Long-term multiparameter assessment of the impact of Hurricane Mara - PowerPoint PPT Presentation

Long-term multiparameter assessment of the impact of Hurricane María on colony measures: A case study in gentle Africanized honey bees (gAHB) from Puerto Rico José L. Agosto-Rivera Department of Biology University of Puerto Rico, Río Piedras Campus

Hurricane Maria’s Trajectory https://www.vox.com/science-and-health/2017/9/21/16345176/hurricane-maria-2017-puerto-rico-san-juan-meteorology-wind-rain-power

Maria’s impact on vegetation ● severe damage to 23-31 million trees. 1 day before hurricane 1 day after hurricane

Vegetation Recovery After Hurricane María in Puerto Rico (Feng et.al., 2018) (Hu & Smith, 2018)

How were honey bees affected by the loss of vegetation due to the hurricane?

Impact of Maria on honey bees

How did the colonies that survived perform with the dramatic loss in vegetation? Did our management strategies worked?

Experimental Set Up and Timeline Monthly or Hive with biweekly camera and photos sensors Data Collection Timeline Oct. Protein feeding Dec. Feb. Aug. 2017 Started: Nov. 3 2017 2018 2017 Jan. Sugar Feeding Nov. Sept. 2018 Started: Sept. 19 2017 2017 Hurricane Maria Sept. 20

Dependent variables and types of data recorded Video Camera Photos Foraging trips Brood • • (entries and exits) Honey stores • Pollen Stores • Pollen carrying bees • (Pollen foragers) Sensors Thermoregulation • Fanning bees • Humidity control • (not analyzed yet)

Automated foraging trips detection using machine learning and pose detection - Exits - Entries - Body length (drones and queen) Rodriguez et al., 2018

Automatic detection of pollen foragers using machine learning Pollen foragers Non-Pollen foragers Rodriguez et al., 2018

Quantifying brood, pollen and honey stores from photos Sealed brood Larvae Pupae Pollen Honey Frame with brood and pollen Zoom in of quadrant with brood and pollen

Results

Brood pattern after María parallels the impact of vegetation: a decrease in October and recovery by November 12000 Brood Levels 11000 10000 Sealed Brood 9000 Pupae Larva 8000 Number of cells 7000 6000 5000 4000 3000 2000 1000 0 g t t v c n b p c o e u a e e O D A N J F S Sugar Feeding Time (months) Protein Feeding

October’s decrease in brood is not related to depletion of honey stores 6000 Honey Levels 12000 Brood Levels 5500 11000 5000 10000 Sealed Brood 4500 9000 Pupae Larva 4000 8000 Number of cells Number of cells 3500 7000 3000 6000 2500 5000 4000 2000 1500 3000 1000 2000 1000 500 0 0 g t t v c n b g t t v c n b p c o e p c u a e u o e a e e O D e O A N J F D A N J F S S Time (months) Time (months)

October’s decrease in brood is not related to depletion of Pollen stores 12000 800 Pollen Stores Brood Levels 11000 700 10000 Sealed Brood 9000 600 Pupae Larva 8000 Number of cells Number of cells 500 7000 400 6000 5000 300 4000 200 3000 2000 100 1000 0 0 g t t v c n b g t t v c n b p c p c o e u o e a e u a e O e O e D A N D J F A N J F S S Time (months) Time (months)

Pollen entry was normalized by November and January’s increase in brood coincides with increased pollen entry 12000 20000 Brood Levels Pollen Foragers 11000 18000 10000 Sealed Brood 16000 9000 Pupae 14000 Larva Number of bees 8000 Number of cells (bees /day) 12000 7000 6000 10000 5000 8000 ? 4000 6000 1 day before hurricane 1 day after hurricane 3000 4000 2000 2000 1000 0 0 Aug Sept Oct Nov Dec Jan Feb g t t v c n b p c o e u a e e O D A N J F S Time (months) Time (months)

Minor or no changes in non-pollen foragers before and after hurricane Maria 12000 30000 Non-Pollen Foragers Brood Levels 11000 10000 25000 Sealed Brood 9000 Pupae Larva Number of bees 8000 Number of cells 20000 (bees /day) 7000 6000 15000 5000 4000 10000 3000 2000 5000 1000 0 0 g t t v c n b p c g t t v c n b o e u a e p c e O u o e a e D A N J F e O D S A N J F S Time (months) Time (months)

Summary and Conclusions ● Brood pattern after María parallels the impact of vegetation: a decrease in October and recovery by November. ● Whether the January increased in brood and pollen entry correspond to increased flora is unknown. ● Sugar feeding may support survival of adults but is not sufficient to support brood rearing (decrease in October despite high honey levels). ● Brood levels were already increased by the time we started the protein administration and no further increase was observed afterwards. Administration was probably too late to see effects. ● The decrease in pollen entry may have serve as a signal to reduced brood levels by one of the following mechanisms: ○ Reducing egg laying of the queen ○ Triggering a behavioral switch from feeding to cannibalism in nurses

Hypothetical model of the impact of Hurricane Maria Pollen Nurse Flora Forager bees Pollen Stored Cannibalism https://www.knowablemagazine.org/article/sustainability/2017/whole-food-diet-bees

Acknowledgements Mentors Undergraduate Students • Dr. José L. Agosto Rivera • Iván Collado • Dr. Tugrul Giray • José Mayoral • Dr. Rémi Mégret • Luis González • Dr. Edgar Acuña • Vivian Reyes • Dr. Patricia Ordoñez • Johann Santos • Dr. Manuel Giannoni- • Luis Santos Guzmán • Nashalis Martínez Graduate Students • Paola Santiago • Isada Claudio Ford • Patricia Arsuaga • Iván Rodriguez • Rochelly Rivera • Stephanie Feliciano • Isabel Rivera • Jonathan Alemán-Ríos • Ashley Lozada Technician • Esteban Morales • Tilden Aponte • Isabel Rivera Undergraduate Students • Ashley Lozada • Janpierre Alemán • Esteban Morales • Claudia Cordero • Nicole Gerena • Airined Montes • Jhoniel Perez • Carlos Morfi This work was supported by the following NSF • Vilnery Rivera • Courtney George Grants: 1707355, 1633184, 1560389, 1545803, • Stephanie González • Edwin Caraballo 1736019, 1826729 • Cheryl Gonzalez • Edwin Rivera • Giancarlo Piovanetti