Flight trials
We collected eastern monarchs (n=32; 17 male, 15 female) from migratory stopover site St. Marks, FL in October 2016 and western monarchs (n=31; 16 male, 15 female) from an overwintering site near Oceano, CA in December 2016 to perform flight trials. All butterflies were housed in overwintering-like conditions in an incubator to ensure they were in the same overwintering state during flight trials in December 2016. We used two flight mills as described in Bradley et al 2005 (Bradley & Altizer, 2005) and an ASCO PS-2000 datalogger (Pasco Scientific, Roseville, CA, USA) to allow eastern and western monarchs to fly in continuous circles of 4.27 m circumference. We recorded the time elapsed between each rotation (to measure instantaneous speed), the cumulative flight time, and the body mass of the monarch pre- and post- flight trial.
Flight trials were performed in a laboratory space maintained at 25°C and controlled for light conditions. One day prior to flight trials, monarchs were removed in groups and a steel wire attachment (32 gauge, 9 cm long) was glued to the dorsal side of the thorax using rubber cement. Following wire attachment, monarchs were held in cylindrical mesh flight cages (diameter= 0.38 m, height= 0.56 m) to allow acclimation to the wire and for free feeding. We calculated five measures of flight performance: flight duration, distance, loss of body mass relative to total distance flown, power, and speed. During the flight trials, we allowed monarchs to fly for 30 minutes with trials ending prior to the 30 minute maximum if monarchs suspended flight for more than 10 seconds on 3 separate occasions. Flight trials were considered unsuccessful if the monarch refused to fly at least one full rotation on the flight mill. We then measured flight performance for the 29 western monarchs (14 male, 15 female) and 27 eastern monarchs (14 male, 13 female) which successfully completed flight trials. Distance was measured as total distance in meters of a flight trial. Loss of body mass was calculated as the change in body mass (massinitial - massfinal) divided by distance flown (in m), then log-transformed. Power was calculated as (1/2*mass*velocity2) divided by time (in s). Speed (m/s) was averaged across 2-minute intervals for the duration of each flight in order to calculate the average flight speed.
We also measured morphological traits relevant to flight, including wing size and wing shape. We measured these traits to determine if any differences in flight behavior were due to differences in wing morphology. Following existing protocols (Altizer & Davis, 2010; Li, Pierce & de Roode, 2016), forewings were scanned on a flatbed scanner and the Fovea Pro plugin (Reindeer Graphics, Inc., Asheville, NC) for Adobe Photoshop was used to measure forewing area, length, breadth and perimeter. From these measurements we calculated aspect ratio, by dividing length by breadth of the forewing, and roundness, by using the equation 4*π*area/ (perimeter)2 (Altizer & Davis, 2010). Using Principal Component Analysis (PCA), forewing area, length, and width were reduced into one variable (PC1) to measure forewing size, while forewing aspect ratio and roundness were reduced to a second variable (PC2) to measure forewing shape.
We used analysis of variance (ANOVA) in R 3.1.3 (R Development Core Team, 2012) to test for differences in PC1 (wing size) and PC2 (wing shape) between eastern and western monarchs. We used analysis of covariance to test for differences in flight duration, flight distance, loss of body mass relative to total distance flown, flight power, and flight speed between eastern and western monarchs. In these analyses, we included butterfly sex as an additional explanatory variable and included PC1 (wing size) and PC2 (wing shape) as covariates. Significance of terms in analyses of variance and covariance was assessed by model simplification followed by model comparison using the command “anova” (Crawley, 2007). Models were plotted to verify the assumptions of homogeneity of variance and normality of errors (Crawley, 2007).