Green Infrastructure Challenge


Green Infrastructure Challenge

This challenge focuses on contractors, builders, infrastructural companies and local governments. Their activities influence the design and maintenance of our living environment and determine our future landscapes. Bee-friendly green infrastructure and public spaces could make a huge contribution to the recovery of Dutch wild bees and other biodiversity.

The good news is that these organisations and companies want to be involved and are ready to make a positive contribution. However, most contracts and tenders for green infrastructure lack location-specific guidelines leading to standard measures to be taken everywhere. Currently, data needed for tailor-made bee-friendly solutions was fragmented and siloed and knowledge insufficiently accessible. We want to change that! As we speak, we are converting available data, knowledge and models into the public domain.

Designing locally-relevant bee-friendly measures is pretty difficult, in the sense that you need to know which species are likely to inhabit the location, their ecological needs (forage plants and nesting requirements), in combination with all sorts of local conditions such as soil and land use, weather, seasons, climate, pollution and other functions of the location (e.g. for drainage, recreation, etc).


The challenge I want to submit is to make infrastructure and construction projects as bee friendly as possible, and adjust them to regional pollinator requirements. Conversations with contractors, municipalities and provinces make clear that a tool capable of helping them would be welcomed and may be integrated in their decision making processes to make optimal use of opportunities for biodiversity (bees, pollinators, plants) in their projects?

Last year we constructed simple maps (100x100m resolution) and information (bees recorded and predicted, ecological needs of specialist bees, management recommendations) for all municipalities of the Netherlands (see ).

This was our first attempt, but it has been very well received and >40 municipalities have been trained in bee-friendly measures based on this information.

The current challenge builds on that and hopefully can move towards a prototype tool or interface that all parties managing green spaces can use to optimize and locally adapt their bee-friendly management.


Key data for this challenge are bee and plant data (both actual records and predicted occurrence), on food plants and their distribution, on other traits and nesting opportunities for bees in space and time. We need to link this type of data to Green Infrastructure data (municipal green management, business parks, nature reserves, city, etc.) so that for a specific location suggestions can be given on the type of measures (for which bees and with which food plants and ecological conditions) that can be taken to improve its ecological fit in the broader ‘natural’ landscape.

  • European Invertebrate Survey (EIS) database
  • Trend Analysis of all bees species (EIS
  • Crypta database: interactions on plants and bees
  • Phylogenetic database on relationships between different species (
  • Trait database
  • Relevant bee models for Dutch bees (spp richness per 100x100m)
  • Bee/butterfly models in relationship with plant data
  • Interaction database (o.a. EIS): plant relevance for bees
  • Global Biodiversity Information Facility (GBIF)
  • NDFF Data
  • other Environmental Data

💾 Data usage and troubleshooting 🛠
Who is Who at the Big Data for Bees Hackathon
About the Big Data for Bees Hackathon

We are currently doing an exploratory data analysis. So far we have found:

  1. An aggregated list of frequency of occurrence of bee species (100x100m)
  2. An aggregated list of frequency of occurrence of plant species (100x100m)
  3. A relational database from observations about which bees are attracted to which plants

It seems that we can use this data to make a plot on a map that shows how ‘rich’ the environment is currently for each species. Some features like selecting by species and calculating which plants are missing in an area could be implemented.


Sounds great, thanks for the update!

Have a look at the LARCH model that does, from my very basic understanding, something similar.

We can connect you with the researcher (Michiel van Eupen) responsible for the model.


Hello, how is it going? Ready to pitch to the jury? :slight_smile:


Bee Support Pitch