Pollinators are responsible for the reproduction of approximately 87 percent of flowering plant species and the production of roughly 35 percent of global food crops by volume. Yet pollinator populations are in steep decline. The Xerces Society estimates that more than one-quarter of North American bumble bee species are at risk of extinction, and wild bee populations in agricultural landscapes have declined by an estimated 23 percent between 2008 and 2020.
Urban areas are increasingly recognized as potential refugia for pollinators, provided they contain sufficient floral resources and nesting habitat. For three years, TerraFuture has been conducting systematic pollinator monitoring across our network of 42 community gardens to quantify whether these spaces genuinely support pollinator biodiversity or whether that is merely an assumption.
The data is clear: community gardens are functioning as significant pollinator habitat, and the effect is larger than we anticipated.
Study Design
Beginning in 2022, trained volunteers and TerraFuture staff conducted standardized pollinator surveys at all 42 community gardens in our network, plus 15 control sites in conventional urban landscapes including mowed lawns, parking strips, and ornamental plantings. Surveys followed the Xerces Society's Pacific Northwest Bumble Bee Atlas protocol, supplemented with pan trap sampling for smaller bee species.
Each site was surveyed four times per growing season, in April, June, August, and September, timed to capture early, peak, and late-season pollinator activity. Over three years, this produced 684 survey events and approximately 2,050 hours of observation and specimen processing.
We identified specimens to species level where possible, using reference collections at the Oregon State Arthropod Collection and consultation with taxonomic specialists. Voucher specimens were deposited at Portland State University's natural history collection.
Key Findings
Across the three-year study period, we documented 127 bee species in our community garden network, representing approximately 28 percent of the estimated 450 native bee species in the Willamette Valley. Our control sites in conventional urban landscapes supported just 35 species, meaning community gardens harbored 3.6 times the species richness.
Average bee abundance, measured as individuals per survey hour, was 4.2 times higher in gardens than in control sites. The difference was most pronounced during midsummer surveys, when gardens provide floral resources that have largely senesced in conventional landscapes.
Fourteen species documented in our gardens are classified as species of concern by the Oregon Biodiversity Information Center, including two bumble bee species, Bombus occidentalis and Bombus fervidus, that have experienced range-wide declines exceeding 50 percent. The presence of these species in urban gardens suggests that even fragmented urban habitat can support at-risk populations.
We tend to think of biodiversity conservation as something that happens in wilderness areas. These data show that a community garden in Southeast Portland can harbor more native bee species than a thousand-acre monoculture farm.
What Makes Gardens Effective
Not all gardens support pollinators equally. Our analysis identified three garden characteristics that are the strongest predictors of pollinator diversity.
Floral diversity is the most important factor. Gardens with more than 30 flowering plant species present during the growing season supported 2.1 times more bee species than gardens with fewer than 15 flowering species. The key is continuous bloom from March through October, ensuring floral resources are available throughout the entire pollinator activity season.
Nesting habitat availability is the second factor. Gardens that include undisturbed ground areas with bare soil or sparse vegetation support ground-nesting bees, which constitute approximately 70 percent of native bee species. Gardens with dedicated nesting areas supported 44 percent more species than those without.
Pesticide-free management is the third factor. All TerraFuture gardens prohibit synthetic pesticide use, and our monitoring data shows no detectable neonicotinoid residues in any garden soil samples, compared to detectable levels in 6 of 15 control sites. The absence of pesticides is a minimum requirement for pollinator habitat, not a bonus feature.
The Network Effect
Individual gardens matter, but the network effect is where the real ecological value emerges. Our connectivity analysis shows that when gardens are spaced within 500 meters of each other, which is within the typical foraging range of most native bees, species richness at each garden increases by an average of 18 percent compared to isolated gardens at similar distances.
This finding has direct implications for garden siting. TerraFuture's expansion plan for 15 new garden sites prioritizes locations that fill connectivity gaps in our existing network, creating a functional habitat mosaic across Southeast and East Portland.
Community Science and Engagement
The pollinator monitoring program engages 52 trained community scientists who conduct the field surveys alongside TerraFuture staff. Post-program surveys show that participation in pollinator monitoring increases gardeners' likelihood of planting pollinator-supportive species by 67 percent and decreases any pesticide use by 48 percent.
We have published the complete three-year dataset on our open data portal and submitted the findings for peer review. The data is also being integrated into the Portland Pollinator Action Plan being developed by the Bureau of Environmental Services. Urban gardens are not a substitute for large-scale habitat conservation. But they are a proven, community-managed, and scalable contribution to pollinator survival in landscapes where every patch of habitat counts.
