Wet meadow restoration buffers the impact of climate change: pollinator resilience during the California drought

Background/Question/Methods

In light of recent global declines in pollinating species, understanding how degraded habitat and intensive restoration practices affect pollinator persistence in the face of climate change is essential.  Can pond-and-plug restoration buffer the effects of drought on bee and forb populations in wet meadows? In these restoration projects, the primary goal is to increase both floodplain connectivity with the water table and seasonal water retention.  This study analyzes community structure and population dynamics of pollinators across the last four years of California’s extreme precipitation.

Study meadows were classified as restored, degraded, or remnant based on the status of floodplain connection.  Twenty-two subalpine meadows in the Sierra Nevada were surveyed annually from 2011 to 2014. Partial-Mantel tests were used to determine spatial independence, with all sites at minimum located 2 km apart. The data spans years of both extreme wet and historic drought.  Flowering plant and pollinator communities were surveyed along observant transects in a 500 m x 1.5 m belt every ten days during peak bloom. All insects (bees, flies, butterflies) observed on flowers are identified to species and vouchered. Floral diversity and abundance was measured as a covariate using floral survey transects. Insect visitation was directly associated with floral density and diversity.

Results/Conclusions

Successful restoration projects can buffer the detrimental impacts of drought on bee and plant populations. We found that restored meadows support larger populations and higher species diversity than degraded counterparts. Restored meadows also supported more similar pollinator communities regardless of precipitation: both annual species turnover and dissimilarity of the pollinator communities between years was smaller in restored compared to degraded meadows. A bipartite network of interacting plant and pollinator species was constructed for each meadow, and network architecture was described by nestedness, connectivity and evenness. Similarity among pollinator communities is visualized using nonmetric multidimensional scaling (NMDS), and a multi-response permutation procedure (MRPP) tests for dissimilarity of meadows between years.

Understanding the factors that affect non-target species community composition can guide the restoration process and help predict long-term outcomes.