Global Patterns and Drivers of Bee Diversity and Endemism on Islands

This study presents the first global checklist of island bees, revealing that despite covering only 5% of Earth's land area, islands support 20% of global bee diversity with high endemism driven by area, isolation, and biome type, thereby highlighting the critical need to integrate these insects into conservation planning.

The Gist

Imagine the world's islands as a giant, scattered archipelago of "nature's laboratories." For a long time, scientists have been studying the plants and big animals (like birds and lizards) in these labs to understand how life evolves. But they've largely ignored the tiny, buzzing workers: bees.

This paper is like turning on the lights in a dark room to finally see the bees. The authors created the first global "phone book" for island bees, listing over 4,000 species across 306 different islands. Here is what they found, explained simply:

1. The "Tiny" Giants

Even though islands make up only about 5% of the Earth's land (think of them as a few crumbs on a giant cake), they hold 20% of all the world's bee species.

• The Analogy: It's like finding that a single small apartment building holds 20% of the world's population.
• The Surprise: Almost half of these island bees are endemics—meaning they are "island-only" residents that exist nowhere else on Earth. They are the unique, local celebrities of the insect world.

2. The Two Different "Beach Parties"

The paper discovered that bees have two very different favorite party spots, depending on whether you care about how many bees there are or how unique they are.

• Party A: The Mid-Latitude "Mediterranean" Islands (High Numbers)

  • Where: Places like Sicily, Cyprus, and the Canary Islands.
  • The Vibe: These islands have a climate with dry summers and wet winters (like California or Southern Europe).
  • The Result: These places are crowded with bees. They have the highest total number of species.
  • Why? Think of these islands as a bustling city with many different neighborhoods. The changing seasons create a "shift work" schedule for flowers, allowing many different types of bees to coexist without fighting over the same food at the same time. However, because these islands are close to the mainland, many of the bees are just regular tourists from the nearby continent, not unique locals.

• Party B: The Tropical Islands (High Uniqueness)

  • Where: Places like Madagascar, New Guinea, and Hawaii.
  • The Vibe: Warm, wet, and tropical.
  • The Result: These islands have fewer total bees than the Mediterranean ones, but they are packed with rare, one-of-a-kind species.
  • Why? These islands are like isolated islands in a sea of water (literally and figuratively). Bees that got stuck there millions of years ago couldn't leave, and new ones couldn't easily get in. Over time, they evolved into strange, new forms.
  • The Hawaii Example: In Hawaii, almost all the native bees belong to just one family tree (a group called Hylaeus). It's as if a single family moved to an island 5 million years ago, and every single person on the island today is a distant cousin of that original family, but they all look and act slightly different now.

3. The Size and Distance Rules

The study confirmed some classic rules of island life, but with a bee twist:

• Bigger is Better: Just like on the mainland, bigger islands have more bees. But the rule is strongest in those Mediterranean-style climates.
• Farther is Harder: The further an island is from the mainland, the fewer bees it has. It's hard for bees to fly across the ocean.
• The "Desert" Exception: Surprisingly, in desert islands, size doesn't matter much. If there's no water or flowers, a huge island won't have many bees, no matter how big it is.

4. The Flower Mismatch

You might think: "More flowers = More bees." It makes sense, right?

• The Twist: The paper found that this isn't always true.
• The Mediterranean Magic: In Mediterranean climates, bees are super-efficient. They have 5 times more bees than you would expect just by counting the flowers. The seasonal changes make them work harder and specialize more.
• The Tropical Trap: In the tropics, despite having tons of flowers, there are actually fewer bees than expected. The constant rain and lack of seasonal change might make it harder for bees to specialize and evolve into new species.

Why Should You Care?

This paper is a wake-up call for conservationists.

• The Problem: We usually protect "biodiversity hotspots" based on how many species are there. But for bees, the places with the most species (Mediterranean islands) aren't necessarily the places with the most unique species (Tropical islands).
• The Risk: If we only protect the crowded islands, we might accidentally let the unique, one-of-a-kind bees of the tropics go extinct without anyone noticing.
• The Solution: We need to treat these tropical islands as "evolutionary reservoirs." They are like the world's most exclusive art galleries, holding unique masterpieces that can never be replaced.

In a nutshell: Islands are the secret factories of bee evolution. While some islands are busy markets full of common bees, others are quiet, isolated workshops creating rare, unique species. To save the future of our pollinators, we need to protect both the markets and the workshops.

Technical Summary

1. Problem Statement

Islands are critical reservoirs of global biodiversity and evolutionary distinctiveness, yet insects—specifically bees, which are vital pollinators—remain severely underrepresented in island biogeography research. While vertebrates and plants have been extensively studied, systematic global syntheses of insect diversity and endemism across island systems are virtually absent. This gap is driven by logistical and taxonomic challenges in assembling standardized data across fragmented political and geographic regions. Consequently, conservation assessments (e.g., IUCN Red List) lack sufficient data on island bees, failing to reflect their high vulnerability and disproportionate contribution to global diversity. The study addresses the need to quantify native bee diversity on islands, identify the drivers of richness and endemism, and understand how these patterns relate to plant diversity and island biogeography theory.

2. Methodology

The authors constructed the first global checklist of native bees on islands using a multi-step data integration and statistical modeling approach:

• Data Compilation:

  • Checklist Assembly: Aggregated data from Discover Life, taxonomic revisions, primary species descriptions, museum records, and validated iNaturalist observations (up to Dec 2024).
  • Spatial Filtering: Matched species records to a global database of islands (Sayre et al.) with an area >1 km². The final dataset included 306 islands representing 4,140 native bee species.
  • Classification: Species were classified as Native, Introduced (e.g., Apis mellifera), Single Island Endemics (SIE), and Archipelago Endemics (AE).
  • Environmental Data: Integrated island attributes (area, isolation distance to mainland, geological origin: oceanic/continental/fragment) and biome classifications (Mediterranean, Tropical, Temperate, Desert, Boreal) from the GIFT database and WWF ecoregions. Angiosperm richness data was sourced from Schrader et al. (GIFT).
  • Sampling Effort Correction: Calculated sampling coverage using the Chao & Jost estimator (via the iNEXT R package) to control for uneven sampling intensity across regions.

• Statistical Analysis:

  • Richness Drivers: Employed Linear Mixed-Effects Models (LMM) with archipelago identity as a random effect. Response variables were log-transformed native species richness. Predictors included log-transformed area, isolation, biome, island type, and sampling coverage. Model selection was based on AICc.
  • Endemism Drivers: Similar LMMs were fitted for SIE and AE richness, testing interactions between area and biome.
  • Bee-Plant Coupling: Fitted linear regressions to test the correlation between native bee richness and angiosperm richness, analyzing residuals to identify biome-specific deviations.
  • Comparisons: Used likelihood ratio tests and Kruskal-Wallis tests to compare slopes and residual distributions across biomes.

3. Key Contributions

• First Global Baseline: Established the first comprehensive global dataset of island bee diversity, covering 306 islands and 4,140 species.
• Quantification of Island Contribution: Demonstrated that despite comprising only 5% of Earth's land area, islands support **20% of global bee diversity**, with 43% of insular species being endemic (accounting for ~8% of all known bee species).
• Decoupling of Richness and Endemism: Provided empirical evidence that species richness and endemism are driven by different mechanisms and spatial patterns on islands, challenging the assumption that richness hotspots are synonymous with endemism hotspots.
• Biome-Specific Dynamics: Revealed that the species-area relationship (SAR) is not uniform but varies significantly by biome, with Mediterranean-type systems showing the steepest slopes.

4. Key Results

A. Global Distribution Patterns

• Richness: Peak native bee richness occurs at mid-latitudes (Northern Hemisphere), mirroring continental patterns. The most species-rich islands are in Mediterranean-type biomes (e.g., Sicily: 636 spp., Cyprus: 384 spp.).
• Endemism: In contrast, endemism peaks in large tropical islands. Major hotspots include Madagascar (88% endemic), New Guinea (73% endemic), and the Hawaiian Islands (90% endemic, dominated by the Hylaeus radiation).
• Island Type: Oceanic islands support significantly fewer species than continental or fragment islands but harbor higher proportions of endemics.

B. Drivers of Richness

• Area: Positive correlation with richness. The Species-Area Relationship (SAR) slope varies by biome:
  • Mediterranean: Steepest slope ($\beta \approx 0.50$); a 10-fold area increase yields ~90% more species.
  • Tropical: Intermediate slope ($\beta \approx 0.28$).
  • Temperate: Modest slope ($\beta \approx 0.19$).
  • Desert/Boreal: Very weak slopes ($\beta < 0.05$).
• Isolation: Negative correlation ($\beta = -0.13$); a 10-fold increase in distance from the mainland reduces richness by ~15%. This effect is uniform across biomes.
• Sampling: Sampling effort positively influences estimated richness but does not alter the fundamental drivers (area and isolation).

C. Drivers of Endemism

• Area: The primary driver of Single Island Endemism (SIE) ($\beta = 0.35$). Unlike richness, the area-endemism slope is uniform across biomes.
• Isolation: Weak and non-significant effect on SIE.
• Richness vs. Endemism: There is a partial decoupling. Mediterranean islands have high richness but low endemism, while tropical islands have lower richness but exceptionally high endemism. Including "biome" in models significantly improves the prediction of endemism based on richness alone.

D. Bee-Plant Relationship

• A positive correlation exists between bee and angiosperm richness globally ($R^2 = 0.50$).
• Biome Deviations:
  • Mediterranean & Temperate: Bees are over-represented (5.1x and 2x more species than expected based on plant diversity).
  • Tropical: Bees are under-represented (0.7x expected).
• Implication: Temperature seasonality and temporal resource partitioning in Mediterranean climates drive bee specialization more than absolute floral richness.

5. Significance and Implications

• Conservation Priorities: The study highlights a critical mismatch in conservation planning. Current efforts often focus on richness hotspots (Mediterranean islands), but tropical islands are the true reservoirs of evolutionary uniqueness (endemism) and face higher extinction risks. Conservation strategies must shift to protect these "evolutionary reservoirs" to prevent silent co-extinction cascades in pollinator networks.
• Theoretical Advancement: The findings refine Island Biogeography Theory by demonstrating that biome context fundamentally alters species-area relationships and that richness and endemism are distinct metrics requiring separate management strategies.
• Model Taxon: Bees are validated as an excellent model for studying island biogeography due to their ecological importance, surveyability, and tight coupling with plant communities.
• Future Directions: The authors call for increased molecular phylogenetic studies to resolve evolutionary histories, targeted expeditions to under-sampled regions (e.g., Oceania, Indomalaya), and the integration of trait-based research to understand habitat and dietary specialization on islands.