A comparative dataset on population genetics, traits, and distributions for nineteen Caribbean tree species

This paper presents a comprehensive comparative dataset for 19 Caribbean tree species, integrating high-density SNP genetic data from SLAF-seq with functional traits, geographic distributions, and climatic associations to address the scarcity of genetic information in tropical biodiversity hotspots.

The Gist

Imagine the Caribbean islands as a bustling, ancient library. For centuries, this library has been filled with thousands of unique books (the tree species), but for most of them, the pages are blank. We know the books exist, but we don't know their stories, their family histories, or how they survive the storms.

This paper is like a team of detectives who finally decided to open those blank books and start writing down the stories. They focused on 19 different types of trees (including one palm tree) living in Puerto Rico and its neighbors, like the Dominican Republic and the US Virgin Islands.

Here is the simple breakdown of what they did and why it matters:

1. The Mission: Filling in the Blanks

Genetic diversity is like the "DNA fingerprint" of a species. It tells us how healthy a population is and how well it can handle changes, like climate change or hurricanes.

• The Problem: Scientists have DNA data for many animals and crops, but for tropical trees—especially on islands—they have almost nothing. It's like trying to understand a forest by looking at a map but never seeing the trees up close.
• The Goal: The researchers wanted to create a "genetic ID card" for these 19 tree species to see how they are related, how diverse they are, and where they live.

2. The Toolkit: A High-Tech "Scissors and Scanner"

Since they didn't have a full "instruction manual" (a reference genome) for these trees, they couldn't just read the whole book. Instead, they used a clever technique called SLAF-seq.

• The Analogy: Imagine you have a giant, uncut encyclopedia. You can't read every single word, so you use a special pair of scissors to cut out tiny, specific pages from every book that are likely to be different from one another. Then, you scan those pages.
• The Result: They scanned thousands of tiny genetic markers (called SNPs) for each tree. For some species, they found over 400,000 of these markers! This gave them a high-resolution picture of the trees' genetic makeup.

3. The Fieldwork: A Caribbean Road Trip

Between 2022 and 2024, the team went on a field trip across the islands.

• They collected leaves from 790 individual trees.
• They didn't just stay in one spot; they traveled from the wet, rainy mountains to the dry, sunny coasts.
• They treated the trees like neighbors, checking out their "family trees" in Puerto Rico and then visiting their cousins in the Dominican Republic and the US Virgin Islands to see how the family is spread out.

4. The "Resume" of Each Tree

Genetics is only half the story. To understand why the trees look the way they do, the researchers also compiled a "resume" for each species. They measured:

• Wood Density: Is the tree built like a sturdy oak or a lightweight balsa wood?
• Leaf Thickness & Size: Are the leaves tough and thick to survive the sun, or thin and delicate?
• Height & Seeds: How tall do they get, and how heavy are their seeds? (Heavy seeds might not travel far, while light ones might fly with the wind).
• Climate Preferences: Do they love the rain or the drought?

5. The Findings: What Did They Discover?

When they looked at the data, they found some reassuring news:

• Healthy Families: Most of these trees have healthy genetic diversity. They aren't too inbred (which is bad) and they aren't too mixed up. They are doing okay.
• The "Island Effect": As expected, the trees on the islands are a bit more genetically similar to each other than trees on a massive continent, simply because they are isolated. But the differences between species were much bigger than the differences between islands.
• No Major Red Flags: For the most part, the trees are stable. However, one tree, Cecropia schreberiana, showed a bit more variation than the others, which might mean it has a unique story to tell.

Why Should You Care?

Think of these trees as the pillars holding up the roof of the Caribbean ecosystem. If the pillars are weak (low genetic diversity), the whole roof could collapse when a hurricane or a drought hits.

This dataset is like a backup drive for the future.

• For Conservationists: It helps them decide which trees need extra protection.
• For Scientists: It gives them a starting point to study how trees evolve and adapt.
• For Everyone: It ensures that if we need to restore a forest after a disaster, we know exactly which "seeds" (genetically) will grow best.

In short, this paper is a massive step forward in making sure we don't lose the unique genetic stories of the Caribbean's trees before we even get to read them.

Technical Summary

1. Problem Statement

• Data Deficiency in Tropical Biodiversity: While genetic diversity is a critical component of biodiversity, comprehensive genetic data is lacking for the majority of species, particularly in tropical forests and biodiversity hotspots.
• Island Vulnerability: Tropical islands (specifically the Antilles) are underrepresented in genetic studies despite being fragile ecosystems with high extinction risks. Island species often suffer from lower genetic diversity due to isolation and small population sizes, with global studies indicating a disproportionate loss of genetic diversity (~28%) in island species since the industrial revolution.
• Lack of Reference Genomes: Most focal species in the Caribbean lack reference genomes, making standard genomic approaches difficult and necessitating the use of reduced-representation sequencing (RRS) for non-model organisms.
• Need for Integrated Data: There is a gap in linking genetic structure with functional traits (e.g., wood density, seed mass) and climatic associations to understand evolutionary and ecological dynamics in tropical trees.

2. Methodology

The study employed a multi-faceted approach combining field sampling, high-throughput sequencing, and ecological modeling.

• Study Area & Sampling:

  • Locations: Samples were collected from Puerto Rico (72% of samples), Hispaniola (Dominican Republic), and the US Virgin Islands.
  • Species: 19 tree species (including one palm, Prestoea acuminata) representing diverse life-history strategies and climatic associations.
  • Sample Size: A total of 790 individuals were sampled (565 from Puerto Rico, 105 from Hispaniola, 120 from USVI).
  • Protocol: Leaf samples were air-dried and preserved in silica gel. GPS coordinates and herbarium vouchers were recorded for all samples.

• Sequencing Strategy (SLAF-seq):

  • Technique: Specific-Locus Amplified Fragment sequencing (SLAF-seq), a reduced-representation sequencing method optimized for de novo SNP discovery without a reference genome.
  • Enzymes: Restriction enzymes HaeIII and RsaI were selected based on a pilot study.
  • Platform: Illumina NovaSeq (pair-end 150 bp reads).
  • Target: ~100,000 tags per species with ~10x read depth.

• Bioinformatics & Data Processing:

  • Quality Control: Raw reads were trimmed and filtered using fastp and Trimmomatic (removing adapters, cropping to 120 bp).
  • SNP Calling: The Stacks v2.66 pipeline was used in de novo mode. Parameters were optimized (mismatch $M=5$) to maximize high-quality SNPs.
  • Filtering:
    • Loci retained if present in $\ge$ 80% of individuals per population.
    • SNPs with < 0.8 missing data retained; samples with > 0.5 missing data removed.
    • Pseudo-SNP Removal: The rCNV R package was used to filter out pseudo-SNPs arising from paralogous regions and copy number variants (CNVs) by controlling for population structure effects on heterozygosity.
  • Final Dataset: 661 samples retained across 19 species, with SNP counts ranging from 24,413 to 433,637 per species.

• Ecological & Trait Data Integration:

  • Species Distribution Models (SDMs): Built using Maxent (v3.4.1) with 16,146 occurrence records. Variables included precipitation seasonality, driest month precipitation, hottest month temperature, and geological substrate. Target-group background approach was used to correct sampling bias.
  • Functional Traits: Data compiled for wood density, leaf thickness, specific leaf area (SLA), maximum height, and seed dry mass.

3. Key Contributions

• Novel Dataset: Creation of the first comparative genomic dataset for 19 Caribbean tree species, filling a significant gap in tropical island genetics.
• High-Resolution SNP Data: Generation of species-specific SNP datasets with high marker density (up to ~433k SNPs) suitable for population-level analyses.
• Multi-Dimensional Integration: Uniquely combines genomic data with functional traits and SDM-derived climate associations, enabling complex analyses of genotype-phenotype-environment relationships.
• Open Science: All raw reads are deposited in the European Nucleotide Archive (ENA, PRJEB96042), and trait/SDM data are available on Zenodo.

4. Results

• Genetic Diversity Metrics:
  • Heterozygosity: Observed heterozygosity ($H_o$) ranged from 0.040 to 0.190. Values were comparable to expected heterozygosity ($H_e$), indicating no significant patterns of inbreeding or outbreeding.
  • Inbreeding Coefficient ($F_{IS}$): Values were close to zero (0.001 to 0.056), suggesting random mating within populations.
  • Nucleotide Diversity ($\pi$): Ranged from 0.0009 to 0.0068, consistent with values reported for lowland tropical trees.
• Population Structure:
  • Genetic variation was generally low across populations within species (except Cecropia schreberiana), a pattern consistent with temperate tree species.
  • Differences in genetic diversity were more pronounced between species than between islands within the same species.
• Model Performance:
  • SDMs showed good performance with Area Under the Curve (AUC) values ranging from 0.60 to 0.85 (median 0.74) and omission rates at the 10th percentile between 0.07 and 0.18.

5. Significance

• Conservation Utility: The dataset provides a baseline for monitoring genetic health in vulnerable island ecosystems, which are disproportionately affected by habitat loss and climate change.
• Evolutionary Insights: By linking genetic structure to life-history traits (e.g., dispersal mode, wood density) and environmental gradients, the data facilitates research into how tropical trees adapt to changing climates.
• Resource for Non-Model Species: Demonstrates the efficacy of SLAF-seq for generating high-quality genomic data in the absence of reference genomes, serving as a template for future studies in biodiversity hotspots.
• Policy Guidance: The data supports evidence-based conservation strategies for Caribbean forests, helping to prioritize species and populations at risk of extinction due to low genetic diversity or restricted ranges.