Bumble Bee Abundance and Diversity Increase with Intensity of Tallgrass Prairie Restoration Intervention

A two-year landscape-scale study in Wisconsin demonstrates that assisted tallgrass prairie restoration, particularly when combined with moderate-intensity management like prescribed fire, significantly increases bumble bee abundance, diversity, and the presence of rare species by enhancing floral resources, regardless of the surrounding landscape context.

The Gist

Imagine the American Midwest as a giant, bustling city that was once a vibrant, wild park filled with wildflowers and buzzing bees. Over time, farmers and developers turned most of that park into concrete and cornfields. The bees, like the city's original residents, lost their homes and food, and their numbers started to crash.

This paper is a story about how a group of scientists tried to fix this by turning some of those cornfields back into wild prairies, but they wanted to know: Does just planting seeds help, or do you need to actively manage the land (like using fire) to really save the bees?

Here is the breakdown of their adventure, explained simply:

The Experiment: Three Levels of "Gardening"

The researchers went to 32 different plots of land in Wisconsin that used to be farms. They divided these plots into three groups, like three different levels of gardening intensity:

1. The "Do Nothing" Group (No Intervention): They stopped farming and just let nature take its course. It's like leaving a vacant lot alone and hoping weeds and wildflowers grow back on their own.
2. The "Plant and Forget" Group (Low Intervention): They planted native prairie seeds but then walked away. No extra help, no fire, just seeds in the ground.
3. The "Super Gardener" Group (Moderate Intervention): They planted native seeds and used controlled burns (prescribed fire) to manage the land. Think of this as a gardener who not only plants seeds but also rakes the leaves, pulls the weeds, and uses fire to clear out the old brush to make room for new growth.

The Results: Bees Love the "Super Gardeners"

The scientists spent two years counting bumblebees (the fuzzy, hard-working cousins of honeybees) and measuring how many flowers were blooming. Here is what they found:

• Bees Prefer Help: The "Do Nothing" lots had the fewest bees. The "Plant and Forget" lots had more, but the "Super Gardener" lots had the most bees and the most diverse types of bees.
• The Fire Factor: While the difference between "Plant and Forget" and "Super Gardener" wasn't huge in terms of raw numbers, the types of bees changed. The "Super Gardener" sites had a better mix of species, including some rare and endangered bees (like the Rusty Patched Bumble Bee) that were only found where active management happened.
• The Food Connection: The main reason the bees thrived wasn't magic; it was the buffet. The "Super Gardener" sites had three times more flowers than the "Do Nothing" sites. The bees followed the food.
• The Neighborhood Effect: The scientists also looked at the surrounding landscape. They expected that if a site was near a big forest or another natural area, it would have more bees. Surprisingly, it didn't matter. Even if a restored prairie was surrounded by cornfields, if the prairie itself was well-managed and full of flowers, the bees showed up. The local garden mattered more than the neighborhood.

The Big Takeaway: Active Care Wins

Think of restoring a prairie like renovating a house.

• Passive Restoration (No Intervention) is like leaving a house empty and hoping the paint dries and the roof fixes itself. It happens, but it's slow and messy.
• Active Restoration (Seeding + Fire) is like hiring a contractor to fix the roof, paint the walls, and install new windows.

The study shows that active care works best. Simply planting seeds is a great start, but adding management tools like controlled fire creates a richer, more diverse environment that supports not just the common bees, but the rare, endangered ones too.

Why This Matters

This is good news for conservation. It tells land managers and farmers that they don't need to worry that using fire will hurt the bees. In fact, fire is a tool that helps create a better home for them. By actively managing these restored lands, we can turn barren fields back into buzzing, blooming sanctuaries that support the insects essential for our food supply and ecosystems.

In short: If you want to save the bees, don't just plant a garden and walk away. Tend to it, prune it, and yes, even burn it occasionally. Your fuzzy neighbors will thank you.

Technical Summary

1. Problem Statement

Human land-use intensification and habitat loss are primary drivers of insect pollinator declines, particularly affecting wild bees. While habitat restoration is a proposed solution, there is a lack of landscape-scale evaluations regarding how different intensities of restoration and management (specifically in tallgrass prairies) affect wild bee populations.

• Context: The tallgrass prairie ecosystem in the U.S. Midwest has been reduced to 3–4% of its original range.
• Knowledge Gap: It is unclear whether "passive" restoration (leaving land fallow) is sufficient, or if "active" interventions (seeding and prescribed fire) yield significantly better outcomes for bumble bee (Bombus spp.) abundance, diversity, and community composition.
• Specific Focus: The study focuses on the federally endangered Rusty-Patched bumble bee (Bombus affinis) and the broader bumble bee community in Southern Wisconsin.

2. Methodology

The study employed a two-year, landscape-scale observational design across 32 sites in Southern Wisconsin, USA. All sites were former agricultural lands managed under the USDA Natural Resources Conservation Service (NRCS) Wetland Reserve Easement (WRE) program.

Experimental Design & Site Categorization:
Sites were categorized into three levels of restoration intervention intensity:

1. No Intervention (Unassisted Recovery): $n=6$. Upland grasslands left to naturally regenerate after agricultural cessation.
2. Low Intervention: $n=13$. Seeded with native prairie plants but not subsequently managed with fire.
3. Moderate Intervention: $n=13$. Seeded with native plants and periodically managed with prescribed fire.
   Note: No "High Intervention" sites (involving soil amendments, repeated seeding, or invasive removal) were included.

Data Collection:

• Bumble Bee Sampling: Conducted from July–September (2018) and June–August (2019).
  • Protocol: Non-lethal catch-and-release hand-netting based on USFWS protocols for B. affinis.
  • Methods: Two survey types per visit: (1) Transect surveys (three 90m transects, 30 person-minutes total) to estimate abundance; (2) Meandering surveys (25 person-minutes) targeting high floral density patches to estimate species richness/diversity.
  • Metrics: Abundance (bees/hour/visit) and Diversity (Inverse Simpson's index, $1/D$).
• Floral Resource Estimation: Presence of flowering species recorded in 96 quadrats (4 per 90m transect) per site. Abundance was calculated as the total number of quadrat-level species occurrences.
• Landscape Context: Semi-natural habitat (SNH) percentage calculated within a 1.5-km buffer around each site.

Statistical Analysis:

• Abundance & Diversity: Generalized Linear Mixed Models (GLMMs) using glmmTMB in R. Fixed effects included restoration intensity, floral abundance, and SNH. Random effects included site, visit, and year.
• Community Composition: Permutational Multivariate Analysis of Variance (PERMANOVA) using Bray-Curtis dissimilarity on core species (excluding rare species <5% abundance). Non-metric Multidimensional Scaling (NMDS) was used for visualization.
• Model Selection: Stepwise selection based on AIC; interaction terms involving SNH were removed due to multicollinearity.

3. Key Contributions

• Gradient Analysis: Unlike binary comparisons (restored vs. unrestored), this study quantifies the response of pollinators across a gradient of management intensity (None $\to$ Low $\to$ Moderate).
• Management Specificity: It isolates the effects of prescribed fire as a management tool within restored prairies, a factor often understudied regarding wild bees.
• Endangered Species Focus: The study documents the occurrence of the federally endangered Bombus affinis specifically in relation to active restoration interventions.
• Landscape vs. Local Scale: It provides evidence that local-scale restoration interventions may outweigh broad-scale landscape context (semi-natural habitat) in driving bumble bee responses in this specific agricultural matrix.

4. Key Results

A. Abundance and Diversity

• Intervention Effect: Bumble bee abundance and diversity were significantly higher at assisted restoration sites (Low and Moderate) compared to No Intervention sites.
  • Abundance: Sites with any intervention had ~2.7x higher abundance than unassisted sites (trend, $P=0.060$).
  • Diversity: Diversity was significantly higher at intervention sites ($P=0.040$). Moderate intervention sites had 2.3x higher diversity than no intervention sites.
• Intensity Comparison: While Moderate intervention sites generally showed higher abundance and diversity than Low intervention sites, these differences were not statistically significant after multiple comparison adjustments.
• Floral Driver: Both abundance and diversity were strongly and positively associated with floral abundance ($P < 0.001$), but not with the amount of semi-natural habitat in the surrounding landscape.

B. Community Composition

• Shift in Dominance: Community composition differed significantly across intervention intensities ($P=0.0005$).
  • No Intervention: Dominated by B. impatiens and B. griseocollis (82% of observations).
  • Moderate Intervention: These dominant species became less dominant (76%), allowing for an increase in relative abundance of other species like B. auricomus-pensylvanicus, B. vagans-sandersoni, and B. rufocinctus.
• Rare Species: Endangered and vulnerable species (B. affinis, B. fervidus) were only observed at assisted restoration sites (Low and Moderate). Their occurrence doubled between Low and Moderate intervention sites.

C. Floral Abundance

• Floral abundance increased significantly with intervention intensity.
  • Moderate sites had 3x more flowers than No Intervention sites.
  • Moderate sites had ~74% more flowers than Low intervention sites.
• Landscape Interaction: Floral abundance at No Intervention sites was positively correlated with surrounding SNH, suggesting landscape seed sources matter for passive recovery. However, this landscape effect was absent in actively seeded sites.

5. Significance and Implications

• Conservation Practice: Active restoration (seeding + management) is a highly effective strategy for increasing bumble bee abundance and diversity, outperforming passive recovery.
• Role of Fire: The study suggests that prescribed fire does not negatively impact bumble bees; rather, moderate-intensity management (seeding + fire) supports the highest diversity and the presence of rare species. The lack of significant difference between Low and Moderate intensity may be due to variability in fire regimes (frequency, timing) across sites, suggesting that optimized fire management could yield even greater benefits.
• Endangered Species Recovery: The presence of Bombus affinis exclusively at active sites indicates that targeted prairie restoration is critical for the recovery of this endangered species.
• Management Recommendations:
  1. Active Seeding: Adding seeds of pollinator-preferred plants is crucial, especially in landscapes with low natural floral abundance.
  2. Management Intensity: Maintaining restored sites with regular management (e.g., prescribed fire) is essential to sustain high floral resources and diverse bee communities.
  3. Local vs. Landscape: In highly modified agricultural landscapes, local habitat quality (floral abundance driven by restoration) is a stronger driver of bumble bee populations than the surrounding landscape context.

Conclusion: The paper provides robust evidence that assisted tallgrass prairie restoration, particularly when combined with prescribed fire, effectively counters pollinator declines by increasing floral resources, thereby supporting higher abundance, diversity, and the persistence of rare bumble bee species.