FennoTraits: Dataset of plant functional traits and community composition in northern European flora

This paper introduces FennoTraits, a comprehensive dataset comprising over 155,000 functional trait observations and 42,000 abundance estimates for 373 vascular plant species across 1,235 low-impact sites in Fennoscandia, providing a detailed profile of plant functional traits and community composition in northern European tundra and boreal forest ecosystems.

The Gist

Imagine the vast, cold landscapes of Northern Europe (Finland, Norway, and Sweden) as a giant, living library. For decades, scientists have been trying to read the books in this library to understand how plants survive the harsh cold, how they interact with each other, and how they might react to a warming climate. But the books were scattered, written in different languages, and often missing pages.

FennoTraits is the project that finally organized this library. It is a massive, open-source "encyclopedia" of plant life created by two researchers, Pekka and Julia, who spent nearly a decade (2016–2025) traveling across seven different wilderness areas to collect the data.

Here is a breakdown of what they did, using some everyday analogies:

1. The "Plant ID Card" System

Think of every plant in the forest as a person. To understand how a person lives, you might look at their height, what they eat, and how active they are. In the plant world, these are called functional traits.

The researchers created a giant database containing "ID cards" for 373 different plant species. For each plant, they recorded:

• Size: How tall is it? (Like measuring a person's height).
• Economy: How "expensive" is its leaf to make? (Some leaves are thick and tough like leather; others are thin and delicate like tissue paper).
• Reproduction: Is the plant currently trying to make seeds or berries?
• Color & Shine: They even scanned the leaves to measure how bright or green they are, similar to how a photographer checks the lighting on a photo.

In total, they took over 155,000 measurements. That's like taking a photo of every single leaf on a tree, thousands of times over.

2. The "Seven Wilderness Camps"

The data wasn't collected in just one spot. The researchers set up camp in seven distinct "neighborhoods" ranging from the frozen, treeless tundra (where it's too cold for trees) to the dense, dark pine forests.

• The Analogy: Imagine studying how people live by visiting seven different cities: a snowy mountain village, a swampy wetland, a sunny meadow, and a deep forest. By comparing the "residents" (plants) in all these different neighborhoods, the researchers can see how the environment changes the "personality" of the plants.

3. The "Two-Person Band"

One of the coolest things about this project is that it was done almost entirely by two people.

• The Analogy: Usually, a project this big is like a massive orchestra with hundreds of musicians, where everyone plays slightly differently. FennoTraits is more like a duo. Because the same two people measured every single plant, from the first day to the last, the data is incredibly consistent. There's no "observer bias" (where one person thinks a plant is 10cm tall and another thinks it's 12cm). They are the only two "conductors" in the band, ensuring the music is perfectly in tune.

4. The "Recipe Book" for Nature

The researchers didn't just measure things; they followed a strict, step-by-step recipe to ensure accuracy.

• Fieldwork: They went out with rulers and bags, picking leaves and measuring heights.
• Lab Work: They dried the leaves in an oven (like baking cookies) to weigh them, scanned them with a high-tech scanner (like a flatbed scanner for photos), and used computer code to calculate the exact area of the leaf.
• Quality Control: They acted like strict editors. If a number looked weird (like a leaf weighing more when dry than when wet, which is impossible), they double-checked their notes and fixed it. They used math to find "suspicious" data points that didn't fit the pattern, ensuring the final encyclopedia is trustworthy.

5. Why Does This Matter?

Why do we need a giant list of plant measurements?

• Predicting the Future: The North is warming up faster than almost anywhere else on Earth. This dataset acts as a baseline. If we know how plants currently look and behave, we can spot changes later. Are the plants getting shorter? Are their leaves changing color? This tells us how the ecosystem is reacting to climate change.
• Saving the Planet: Plants are the Earth's lungs and its carbon storage units. By understanding how different plants function, we can better predict how much carbon these northern forests and tundras can store, which is crucial for fighting global warming.
• Open Access: The best part? They are giving this "encyclopedia" away for free. Any scientist, student, or curious person can download the data to build their own models, write their own stories, or solve new puzzles.

In a Nutshell

FennoTraits is a massive, high-quality, and meticulously organized collection of "plant biographies" from the frozen north. It's a tool that helps us understand the rules of nature in a changing world, created by two dedicated explorers who treated every leaf like a precious piece of evidence.

Technical Summary

1. Problem Statement

Northern European ecosystems (tundra and boreal forests) are experiencing rapid warming due to anthropogenic climate change, threatening their resilience, carbon storage capacity, and biodiversity. Understanding how plant functional traits vary across species and environmental gradients is critical for predicting ecosystem responses to these changes. However, existing datasets often lack:

• Spatial and taxonomic comprehensiveness: Few datasets cover the full gradient from boreal forests to oro-Arctic tundra across multiple countries (Finland, Norway, Sweden).
• Methodological consistency: Large-scale studies often involve multiple observers, introducing observer bias and reducing data comparability.
• Granularity: There is a need for high-resolution data that links community composition, intraspecific trait variability, and specific environmental drivers (e.g., microclimate, geodiversity) to support robust ecological modeling and conservation strategies.

2. Methodology

The authors present FennoTraits, a dataset collected between 2016 and 2025 across seven study areas in Fennoscandia.

• Study Areas: Seven distinct regions were selected based on floristic diversity and accessibility: Máttavárri and Rásttigáisá (Norway); Kilpisjärvi, Pallas, Värriö, and Oulanka (Finland); and Vindelfjällen (Sweden). These areas span tundra, meadows, wetlands, shrublands, and boreal forests.
• Sampling Design:
  • The dataset comprises 1,235 study sites across 11 different study designs (e.g., gradient designs, community experiments, seasonal monitoring).
  • Nested Plot Structure: Most designs utilized a nested plot structure (0.2 m², 1.0 m², and 2.0 m radius circles) to capture community composition and height at the plot level, while leaf traits were measured at the site level.
  • Stratification: Sites were selected via random stratification to maximize coverage of environmental gradients (e.g., altitude, canopy cover, solar radiation).
• Data Collection Protocol:
  • Operators: All field and laboratory work was conducted by only two researchers to ensure maximum consistency and minimize observer bias.
  • Taxonomy: Plants were identified to species level (using the Leipzig Plant Catalogue as a backbone), with specific handling for difficult genera (Taraxacum, Alchemilla) and hybrids.
  • Community Composition: Visual estimation of species cover percentages for all vascular plants.
  • Functional Traits (10 traits):
    1. Size-structural: Median and maximum vegetative height (cm).
    2. Reproductive: Sexual reproductive effort (index 0–5).
    3. Leaf Economic: Fresh weight, dry weight, leaf area, Specific Leaf Area (SLA), and Leaf Dry Matter Content (LDMC).
    4. Leaf Spectral: Leaf brightness index (BITM) and Excess Green index (ExG), derived from high-resolution RGB scans.
  • Laboratory Processing: Leaves were scanned at 600 dpi. Leaf area and spectral indices were calculated using custom R scripts involving iterative segmentation, shadow removal, and spectral index formulas (e.g., NDYI, RCC).
• Quality Control: A rigorous, multi-step validation process was applied:
  • Visual inspection of histograms and pairwise trait scatterplots.
  • Application of hard biological bounds (e.g., LDMC must be 0–1, SLA < 200 mm²/mg).
  • Robust Regression Analysis: Huber M-estimation was used to detect outliers based on within-species trait covariance (e.g., relationships between leaf area, wet weight, and dry weight). This allowed the authors to pinpoint specific erroneous measurements (e.g., a bad leaf area scan) rather than discarding entire samples.

3. Key Contributions

• Dataset Scale: The dataset contains 42,049 abundance estimations and 155,794 functional trait observations representing 373 vascular plant species.
• Unique Consistency: The dataset is unique because it was collected and processed by a single pair of researchers, ensuring high comparability across diverse study designs and large spatial gradients.
• High Spatial Accuracy: Sites were georeferenced using high-accuracy GPS (up to centimeter-scale), enabling true resurveys and precise spatial analysis.
• Comprehensive Trait Suite: It integrates standard structural/economic traits with novel spectral indices (BITM, ExG) derived from scanned imagery, bridging ground-based and remote sensing data.
• Open Science: The data is structured into five files (community/heights, leaf traits, species lookup, metadata, and an R script) and is available via GitHub and Zenodo under a CC-BY license.

4. Results and Data Characteristics

• Taxonomic Coverage: The 373 species represent the majority of native vascular plants in the study areas, including rare Arctic species and common boreal dominants.
• Trait Distributions: The data reveals distinct trait profiles across the seven study areas, reflecting the transition from boreal forests to tundra.
• Data Structure:
  • Community/Height Data: Plot-level resolution (1,157 sites).
  • Leaf/Reproductive Data: Site-level resolution (1,039–1,040 sites), with pooled samples from multiple individuals.
  • Metadata: Includes detailed descriptions of study designs, coordinates, and treatment variables (e.g., species removal experiments).
• Validation Outcomes: The quality control process identified and corrected specific errors (e.g., 14 leaf samples with erroneous values were set to missing), ensuring high data integrity while preserving genuine biological variation.

5. Significance

• Ecological Modeling: FennoTraits provides a robust foundation for analyzing intraspecific trait variability (ITV) and its response to environmental gradients, which is crucial for improving species distribution models under climate change.
• Conservation & Restoration: The dataset supports the development of effective conservation strategies and restoration efforts in northern ecosystems by linking functional diversity to ecosystem services like carbon sequestration.
• Methodological Benchmark: The paper serves as a best-practice guide for open, reproducible science in ecology, demonstrating how to standardize data collection across complex, multi-year, multi-site projects.
• Future Research: The high-precision coordinates and nested design enable future resurveys to monitor long-term vegetation changes, while the spectral indices facilitate the calibration of satellite remote sensing data for northern latitudes.

In summary, FennoTraits is a high-quality, standardized, and comprehensive resource that addresses the critical need for reliable trait data to understand and predict the future of northern European flora in a warming world.