In this study, we simulated the extirpation of a hummingbird-pollinated understory plant, Heliconia tortuosa, from tropical forest fragments using a replicated Before-After-Control-Impact (BACI) experimental design while quantifying hummingbird space use, foral visitation rates, and plant pollination success. Each response variable was measured during each experimental period (pre and post) in sites with and without H. tortuosa removal (treatment and control).

These datasets were collected in the landscape surrounding the Las Cruces Biological Research Station in southern Costa Rica (2016–2018).

Details for each dataset are provided in the README file.

Datasets included:

1) Hummingbird captures in mist nets

• Number of hummingbirds captured per capture session
• Recapture rate (proportion of birds captured during 'pre' period that were recaptured during 'post' period)
• Body mass measurements for captured hummingbirds (mass and wing length)

2) Hummingbird radio-telemetry

• Proportion of observation time spent in focal area where bird was originally captured

3) Plant-hummingbird interactions quantified through trail camera observations

• Visitation rates for Heliconia and non-Heliconia flowers (all birds)
• Visitation rates for Heliconia and non-Heliconia flowers (individual marked birds, 2017-2018 only)
• Visit durations

4) Pollination success

• Proportion of flowers (styles) with pollen tubes

5) Site and replicate characteristics (elevation, Heliconia density, and Heliconia calories removed/ha)

6) Data related to percentage of calories removed

• Relative visitation rates to 35 plant species (i.e., weights used to tailor estimate of floral resource availability to hummingbird perspective)
• Estimated percentage of calories removed by experimental Heliconia removal

Missing values are denoted by NA. Additional details are available in the README file.