Quercus Project

The Quercus Project is a research programme on biodiversity and ecology in oak ecosystems, hosted by Linnaeus University with Lund University, the Swedish University of Agricultural Sciences (SLU) and Linköping University as partners. It is led by Prof. Markus Franzén.

Its website is read by peer scientists, funding bodies and a public interested in biodiversity science. It has to carry institutional credibility while staying maintainable by a small research team with no web staff, and accessibility was a requirement from the start.

Site

quercusproject.com

Client

Quercus Project, Linnaeus University

Live

Since July 2026

Platform

Drupal 11 with a Bootstrap 5 subtheme

Accessibility

WCAG 2.1 AA, built into the theme and tested

MADDev products

Ecosystems, species_ext, External Entities

• A Drupal site replacing the programme's WordPress site, sharing its design system with Museo Avellonia
• Sections for research, subprojects, researchers and students, publications, methodologies, a gallery and news
• Landing pages composed from stacked paragraphs, one paragraph type per design section, which keeps editing simple for a small team
• Researcher and student profiles on one user model, with job titles, degree levels and cohort years kept to the people they apply to
• Species pages drawn from external biodiversity data through MADDev's External Entities and species_ext modules
• Accessibility to WCAG 2.1 AA, built into the theme layer rather than retrofitted

The programme's oak-biodiversity survey, 824 species in 30 sets across 31 survey sites, drives MADDev's Ecosystems viewer. It was built for this project and designed from the start as a product other sites can install. It is installed on the live site and not yet open to the public.

The survey arrived as a spreadsheet, and its first import created a second species table next to the one the site already fed from external databases. We reversed that: the survey's study data now lives on the existing species records, so there is one species entity and Manage fields stays the single source of truth.

Two checks made that change safe:

• Of 24 columns that looked computable from others, each was tested against all 824 rows: 15 reproduce exactly and 9 do not. One formula matched 702 rows and drifted by 0.01 on the rest, because the source had rounded its numbers first. A derivation that is right for 85% of rows is not a derivation, so those 9 are stored as delivered, not recomputed.
• Before the duplicate was retired, a field-by-field comparison of 753 shared species, 55,592 values, made without a single live API call, found 9 real differences: 8 additions and 1 deliberate choice of taxonomy source.