Long-distance migration can be seen throughout the animal kingdom and can have large impacts on population dynamics and species distributions. The act of migration itself also affects the evolution of a species, as evolutionary forces select for certain characteristics in animals conducting long-distance migration. Monarch butterflies are best known for their annual migration from Canada and the northern United States to central Mexico, but some populations of monarchs have lost the ability to migrate. Previous research found that migratory monarchs had larger, more elongated wings than their non-migratory counterparts and it was hypothesized that these traits were beneficial for migration. However, Bergmann's rule - which predicts larger body sizes with increasing latitude - could also explain this pattern as migratory populations are found at higher latitudes. To understand the role of wing dimensions in migration, we examined forewing size and shape of migratory and non-migratory monarchs from seven worldwide populations varying in latitude. Results showed that larger forewing size was indeed correlated with migratory status rather than latitude. However, migratory monarchs did not have more elongated forewing shape than non-migratory monarchs across the globe. Our study indicates that size may play a larger role than shape in long-distance migratory capability. Keywords: long-distance migration, Danaus plexippus, wing morphology, Bergmann's rule *Corresponding author Yiwen Li, Emory University, USA, E-mail: email@example.com Amanda A Pierce, Emory University, Department of Biology; University of North Carolina at Chapel Hill, Department of Biology Jacobus C. de Roode, Emory University, Department of Biology © 2016 Yiwen Li, et al., licensee De Gruyter Open. This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivs 3.0 License. Although the annual migration of monarch butterflies is a renowned phenomenon, monarchs have colonized many parts of the world where they have formed nonmigratory populations [18,19]. These populations not only differ in migratory status, but are also genetically differentiated and have been shown to differ in the Collagen IV alpha-1 gene, which is involved in muscle development and efficiency [18-20]. Previous work on wing morphology by Altizer & Davis (2010) indicated that migratory monarchs from eastern and western North America have larger wings than non-migratory South Florida and Hawaii monarchs and larger, more elongated wings than the non-migratory populations from Puerto Rico and Costa Rica. Both wild-caught and lab-reared monarchs supported these findings, which implies that monarch wing morphology is at least partly genetically based. However, there is a second competing hypothesis as to why the examined non-migratory monarchs had smaller wing sizes than their migratory counterparts, as studies have shown environmental factors can also impact these traits. A study on morphological adaptations of the common nightingale revealed that, rather than being associated with migration distance, elongated wings were positively associated with longitude and phenology at breeding destinations . This suggests morphological changes thought to be related to long-distance migration may be driven by environmental factors at destination regions instead. Another study examined phenotypic consequences of a rapid range expansion in the damselfly, Erythromma viridulum, and revealed a positive correlation between latitudinal variation and body size in newly formed populations; this suggests that Bergmann's rule was maintained through dispersal . Bergmann's rule states that in species that are broadly distributed, populations of larger body size tend to be found in colder environments, or increasing latitudes, while species of smaller body size are found in warmer regions, typically closer to the equator [23,24]. Because monarch butterflies have also experienced a recent range expansion to colonize a wide range of latitudes [18,19], it is important to explore the impact of latitudinal variation on monarch body traits in addition to their migratory status . In the study by Altizer & Davis (2010), the butterfly populations with smallest wing sizes also happened to be those closest to the equator. In ectotherms the relationship between morphology and latitude is not solely a result of phenotypic plasticity to temperature, but can be the result of genetic differentiation among populations [25,26]. Thus, the phenotypic and genetic variation in wing morphology found by Altizer & Davis could be at least partly driven by latitudinal variation. Therefore, the goal of this project was to assess whether forewing size and forewing shape differ based on the migratory status of monarch butterflies or whether they vary by latitude instead. If the latter, we would expect monarch populations at lower latitudes to have smaller wings and monarchs at higher latitudes to have larger wings. However, if migration is instead driving this pattern, we would expect wing size and shape variation to be determined by migratory status. We included monarch butterflies previously caught during 2007-2012 that differ in migratory status, and included non-migratory populations from varying latitudes. We examined the forewing size and shape differentiation among migratory monarch butterflies from eastern and western North America, and non-migratory monarchs from South Florida, Hawaii, Spain, Morocco, and Aruba. 2 Methods 2.1 Geographic sampling locations We used monarchs from eastern North America, western North America, South Florida, Hawaii, Spain, Morocco, and Aruba (Figure 1). Collection years and latitude of sampling site are listed in Table 1. All monarchs were netcaptured as adults by scientists and migratory monarchs were captured during the migration season. Latitudes for eastern and western United States monarchs are not listed because these samples were collected at migratory stopover and overwintering locations, not where monarchs developed and resided during the breeding season. The western overwintering location of Pismo Beach, CA has been shown to receive the most migrants coming from the northern region of the intermountain west . The eastern site in St. Marks, Florida is a stopover location for monarchs traveling along the eastern seaboard . Sample numbers and mean forewing measurements are listed in Table 2. 2.2 Wing measurements We focused on forewings because the previous study by Altizer & Davis (2010) indicated that monarch forewings cover the majority of their hindwings during soaring flying, and thus forewing morphology is more likely to contribute towards the exposed wing surface area. Forewings were scanned and the Fovea Pro plugin (Reindeer Graphics, Inc., Asheville, NC) for Adobe Photoshop was used to measure forewing area, length, breadth and perimeter Figure 1. Sampling locations of seven migratory and nonmigratory monarch butterfly populations. Sample sizes are within parentheses. Table 1. Sampling data information: location, year and latitude. Sampling location western US eastern US South Florida Aruba Hawaii Spain Morocco Sampling city/region Pismo Beach, CA St. Marks, FL Miami Noord Hawaii; Maui; Kauai Cordoba Belyounech Migratory Status Migratory Migratory Non-migratory Non-migratory Non-migratory Non-migratory Non-migratory Year collected Feb. 2008 Oct. 2007 April 2011 Jan. 2012 Nov. 2009 Sept. 2012 Oct. 2012 25.7753 12.5667 21.3114 37.8833 35.1714 Latitude (Figure 2). Forewing area in millimeter2 was measured of the total forewing surface region, length in millimeters was measured from the distal tip to the thorax attachment point of the forewing, breadth in millimeters was perpendicular to the length, and perimeter was measured in millimeters for the entire forewing. From this we calculated aspect ratio, by dividing length by breadth of the forewing, and roundness, by using the equation 4**area/ (perimeter)2, as described in Altizer & Davis (2010). Using Principal Component Analysis (PCA), forewing area, length, and width were reduced into one variable to measure forewing size, while forewing aspect ratio and roundness were reduced to a second variable to measure forewing shape. ANOVA using population as the fixed effect variable and Tukey's Test were then performed to examine the forewing size and shape differences among the migratory and nonmigratory populations. We also tested for the effect of Figure 2. Monarch butterfly photograph, indicating forewing measurements: area (grey shading with blue border), length (yellow arrow), breadth (green arrow), and perimeter (red border). Table 2. Mean forewing dimensions. Means are shown separately for males and females (±SE). Sex Male Female Population eastern US western US S. Florida Aruba Hawaii Spain Morocco eastern US western US S. Florida Aruba Hawaii Spain Morocco N 40 36 14 17 10 8 14 20 23 6 9 14 8 10 Length (mm) 51.946±0.257 51.003±0.255 48.618±0.659 44.414±1.132 47.574±0.880 48.474±0.426 48.242±0.986 50.633±0.484 51.174±0.390 48.477±1.713 44.414±0.600 48.043±0.602 47.229±1.221 48.287±0.925 Breadth (mm) 26.358±0.162 25.868±0.141 24.868±0.308 23.520±0.582 24.333 ±0.428 24.827±0.311 23.977±0.586 25.735±0.283 26.070±0.188 24.964±1.001 23.208±0.323 24.536±0.329 24.067±0.562 24.374±0.481 Aspect Ratio 1.972±0.008 1.972±0.007 1.955±0.014 1.912±0.012 1.955±0.015 1.954±0.019 2.016±0.085 1.969±0.008 1.963±0.009 1.941±0.015 1.914±0.011 1.959±0.041 1.962±0.019 1.982±0.011 Area (mm2) 865.220±8.828 837.908±8.129 775.912±18.911 682.712±32.030 737.184±23.345 776.952±16.027 744.276±30.472 826.185±16.662 844.022±11.086 788.981±54.473 665.206±17.078 760.131±17.541 737.622±33.177 756.003±27.497 Roundness 0.663 ±0.002 0.664±0.002 0.661±0.004 0.686±0.003 0.661±0.006 0.680±0.004 0.662±0.005 0.665±0.003 0.665±0.003 0.675±0.004 0.681±0.003 0.672±0.003 0.676±0.007 0.667±0.002 migratory vs. non-migratory type using model comparison and a linear mixed effects model with migratory status as a fixed effect and population as a random effect using the lme4 package . Finally, to test whether forewing size was correlated with latitude, we performed a correlation analysis for non-migratory monarch populations using the rcorr function in the Hmisc package . All statistical analyses were performed in R 3.1.2 . Each analysis was performed separately for males and females and corrected for multiple comparisons using a Bonferroni adjustment (p=0.025). 3 Results 3.1 Wing morphology Migratory monarchs from eastern and western North America had significantly larger wings, but their wings were not consistently more elongated than those of nonmigratory monarchs (Figure 3). ANOVA and Tukey's test confirmed the largest difference in forewing size was between North American migratory populations (US-east and US-west) and non-migratory Aruba monarchs for both males (US-east vs. Aruba, p<0.01; US-west vs. Aruba, p<0.01) and females (US-east vs. Aruba, p<0.01; US-west vs. Aruba, p<0.01). Pairwise comparisons among populations for male and female wing size are shown in Table 3. Overall, migratory status was found to affect wing size in males (2(1)=8.67, p=0.003) and females ( 2(1)=8.88, p=0.003). ANOVA and Tukey's test also showed significant differences in forewing shape between North American migratory populations and Aruba non-migratory monarchs for both males (US-east vs Aruba, p<0.01; US-west vs Aruba, p<0.01) and females (US-east vs Aruba, p<0.01; US-west vs Aruba, p<0.01). In all of these cases, migratory monarchs had more elongated wings than the non-migratory Aruba monarchs. Additionally, with respect to forewing shape, male Aruba monarchs had significantly less elongated forewings than males in Morocco and S. Florida (p<0.015) while female Aruba monarchs had significantly less elongated forewings than females in Morocco (p<0.01). While migratory populations were found to have more elongated wings than the nonmigratory Aruba population, we did not find an overall effect of migratory status on wing shape in either males (2(1)=0.70, p=0.40) or females (2(1)=2.39, p=0.12). 3.2 Latitude In the non-migratory populations, we also tested whether the variation in forewing size was due to latitudinal differences rather than migratory status. We performed a correlation analysis examining male forewing size and population latitude and did not find a significant relationship (r=0.77, p=0.13). Similarly, we found no significant relationship between latitude and female forewing size (r=0.47, p=0.43). We also did not find a significant relationship between latitude and forewing Figure 3. Average male and female forewing size and shape from migratory monarchs (eastern and western North America) and nonmigratory monarchs (South Florida, Hawaii, Spain, Morocco, and Aruba). The y-axes represent the average (±1 s.e.) forewing size (greater PCA-size means larger wings) and forewing shape (greater PCA-shape means more elongated wings) among different monarch populations, as based on principal component analysis. The non-migratory populations are ordered along the x-axis by decreasing latitude. Migratory monarchs (US-east and US-west) had significantly larger forewing size than non-migratory populations (South Florida, Hawaii, Spain, Morocco, and Aruba). In contrast, migratory monarchs did not show a strong pattern of forewing elongation comparing to the non-migratory populations. Aruba monarchs had significantly less elongated forewings than many populations. Table 3. Forewing Size Comparison. Males and females are compared in separated tables. Sizes are considered to be equal if they are not significantly different from one another using an =0.025 cut-off value. Male Female Population East West Morocco Spain S. Florida Hawaii Aruba Population East West Morocco Spain S. Florida Hawaii Aruba East W=E M<E S<E S.F<E H<E A<E East W=E M=E S=E S.F=E H=E A<E West E=W M<W S=W S.F=W H<W A<W West E=W M=W S<W S.F=W H<W A M W>M S=M S.F=M H=M A=M Morocco E=M W=M S=M S.F=M H=M A=M Spain E>S W=S M=S S.F=S H=S A S M=S S.F=S H=S A=S S. Florida E>S.F W=S.F M=S.F S=S.F H=S.F A H W>H M=H S=H S.F=H A=H Hawaii E=H W>H M=H S=H S.F=H A=H Aruba E>A W>A M=A S>A S.F>A H=A Aruba E>A W>A M=A S=A S.F=A H=A shape in non-migratory males (r=0.47, p=0.42) or females (r=0.78, p=0.12). 4 Discussion Our study quantified phenotypic traits that are important in facilitating long-distance migration. To perform long distance migration, migratory monarchs evolved morphological and physiological adaptations to develop more efficient metabolism, perform flight directionality, and enhanced flight endurance [19,32-34]. Analyzing traits linked to migratory status of monarch butterflies provides us with insights into how evolutionary selective pressures favor certain characteristics for long distance animal migration. These morphological changes of migratory traits found in monarchs are also evident and applicable to other species, which experience similar selective pressure during long-distance migration. In birds, for instance, migratory species were found to display greater wing spans to facilitate efficient flight compared to nonmigratory bird populations ; likewise, in dragonflies wing shape significantly varied with migratory status . We determined if differences in wing size and shape were correlated with migratory status or were instead correlated with latitude and thus more consistent with Begmann's rule. Bergmann's rule states that larger body sizes tend to be found at increasing latitudes, while smaller body sizes tend to be located in warmer regions and closer to the equator [23,24]. Our results found no relationship between monarch butterfly forewing size and population latitude and instead found that forewings were significantly larger in migratory versus non-migratory monarchs. This is similar to the Altizer & Davis (2010) study on monarch wing morphology which showed that eastern and western North American migratory populations have larger wings compared to non-migratory populations in South Florida, Hawaii, Puerto Rico and Costa Rica. It would therefore seem that it is indeed migratory status, rather than latitudinal differences, that best explains the differences in forewing size. Aerodynamic theory suggests that greater wing size is crucial for long-distance migration  and a previous study revealed that birds with larger wing size fly faster than birds with smaller wing sizes given the same amount of energy input . This phenomenon of energy conservation with greater wing size is therefore beneficial for long-distance migratory species, as the traveling conditions are less predictable and food sources along the route could be scarce. For migratory monarchs, more efficient flying with less energy being consumed per distance flown could result in fewer stopovers and would provide a quicker and more likely arrival at the final destination. In contrast, non-migratory monarch populations in South Florida, Hawaii, Spain, Morocco, and Aruba are not under strong selective pressure for long-distance travel, and our results confirmed that they developed smaller forewings. Shorter forewings allow for slow and more controlled flight , which may be beneficial for non-migratory, resident monarchs. Our results also indicate that forewing shape does not significantly differentiate eastern and western North America migratory monarchs from the non-migratory monarch populations. Butterflies in general utilize a combination of efficient gliding and active flapping . Theoretically, species with elongated forewings and a larger aspect ratio employ extensive gliding which lowers their energy consumption whereas species using active flapping flight display more compacted forewing shape which allows for increased maneuverability . However, migrating monarchs are frequently exposed to wind drift and are known to perform intermittent gliding and soaring flight among active flight [41,42]. As suggested, more elongated forewings may be beneficial for lower energy consumption and gliding behavior, but may also hinder active flight, which is also an important part of migration. This trade-off may explain the nonsignificant forewing shape differences among migratory and non-migratory monarch populations. While the Altizer and Davis (2010) study found that eastern monarchs, which travel the longest migratory distances, had more elongated forewings than most nonmigratory populations, when looking at wild caught eastern and western migratory monarchs they primarily found a significant difference in wing shape when compared with Costa Rica and Puerto Rico populations. This is similar to our findings in that we did not see a consistent difference in forewing shape based on migration status across all examined populations. While forewing shape may be beneficial to long-distance migration, as is the case in many other migrating species, it appears that forewing size is under greater selection in monarch butterflies. A recent study examining natal origin of western migrating monarchs found a positive relationship between distance to the overwintering location and forewing size . However, the same study did not find a relationship between more elongated wings and greater migratory distance. This is in accord with our study, which suggests that forewing size, rather than shape, is an important contributor to the migratory ability of monarch butterflies. Conversely, it may be the case that changes in wing size rather than wing shape evolve faster in the non-migratory populations, leading to greater differences in size among migratory and non-migratory monarchs. In this scenario, elongated wings may still be beneficial for long-distance migration, but could explain why more rounded wings are not consistently seen in non-migratory populations. Indeed, a study examining migratory eastern North American monarchs found that earlier migrants tend to have more elongated forewings compared to later migrants, which suggests that elongated wings may provide a benefit in long-distance monarch migration . We found that monarchs in Aruba have significantly rounder wings compared to many other sampling locations, including other non-migratory locations like Hawaii and Morocco. Again, Altizer and Davis (2010) found the forewings of monarchs in Costa Rica and Puerto Rico to be less elongated than both migratory groups and another study found that monarchs in Cuba have less elongated wings as well . While these locations are all non-migratory, they also all share a general locality of being in or sharing coastline with the Caribbean Sea. Since other non-migratory populations from Hawaii, S. Florida and Morocco did not exhibit significantly rounded wings, it is possible that rounded wings may instead be correlated with some aspect of the Caribbean, rather than with migration status. Overall, our results suggest that forewing size is a major morphological trait related to monarch migration. As such, identifying the genes underlying wing size, for example through Quantitative Trait Analysis on the F2 offspring of migratory and non-migratory monarchs, would be a fruitful avenue towards characterizing the genetic basis of migration. Acknowledgements: Funding for this project was provided by the National Science Foundation (grant DEB1257160 to J.C.d.R.). We thank E. Sternberg, R. Rarick, C. Gowler, D. Frey, F. Hernández Haeger and M. Maudsley, for help with field collections, and A.K. Davis and two anonymous reviewers for constructive comments that significantly improved the manuscript.
Animal Migration – de Gruyter
Published: Jun 23, 2016