Impacts of Prevailing Winds on Migratory Patterns of Danaus plexippus (Monarch Butterflies) Along Overnight Roosts in Texas

Little is known about how and where monarch butterflies choose their roosting sites along their fall migration. To better predict and facilitate habitat conservation, this thesis attempted to investigate the role of prevailing winds in monarch butterfly migratory patterns in Texas during the fall migration periods of the years of 2011 to 2016. Using Geographic Information System (GIS) software ArcMap®, comparative analyses were conducted between local wind data and roost data. Hourly observations of wind speed and direction for the six migration periods were obtained from 20 weather stations in Texas. The fall roost data, including coordinate information and butterfly count, was obtained from a citizen science program, Journey North. Daytime (06:00-17:59) wind speeds and directions were selected from the hourly observations, and were categorized into tailwinds, crosswinds, headwinds, light winds (≤ 25km/h), and strong winds (> 25km/h) based on the population’s known responses to different wind conditions. The possible combinations of the wind speed and direction were ranked in the order as follows: the most unfavorable (strong headwinds), unfavorable (strong crosswinds), moderate (strong tailwinds and light tailwinds), favorable (light crosswinds), and the most favorable (light tailwinds) wind condition. Daily wind conditions were graphically rendered on a Texas map by different pixel colors, and then the point features of daily roost observations were overlaid onto the wind condition map so that both the wind condition and roost location could be visually compared together. These daily wind-roost layers were summarized into monthly wind-roost layers to conduct statistical analyses. To see if the migration speed varied under different wind conditions, the latitudinal progress of roost locations was compared to the overall wind condition that occurred in each of the observation periods. To see if crosswinds affected the locations of roosts by pushing the butterflies further east or west in Texas, the longitudinal progress of roost locations were compared to the crosswind condition that occurred in each of the observation periods. Based on the nature of the surface-level low atmospheric pressure, which typically accompanies upward airstreams, follow-up analyses were conducted between barometric pressure data and roost data to see if the pressure gradient affected the migratory pattern of monarch butterflies. Four hypotheses were developed: (1) more butterflies would form their roosting sites where the more favorable wind conditions occur; (2) faster migrations would occur under the more favorable wind conditions, (3) the locations of roosting sites would longitudinally vary under different crosswind components, and (4) the butterflies would migrate from higher to lower atmospheric pressure areas to take advantage of upward airstreams during their migration.