Convergent iridescence and divergent chemical signals in sympatric sister-species of Amazonian butterflies

  1. Joséphine Ledamoisel  Is a corresponding author
  2. Bruno Buatois
  3. Rémi Mauxion
  4. Christine Andraud
  5. Melanie McClure
  6. Vincent Debat
  7. Violaine Llaurens
  1. Collège de France, Centre Interdisciplinaire de Recherche en Biologie, France
  2. Muséum national d'Histoire naturelle, Institut de Systématique, Evolution et Biodiversité (ISYEB UMR 7205), France
  3. Smithsonian Tropical Research Institute, Panama
  4. Muséum national d'Histoire naturelle, Centre de Recherche sur la Conservation, France
  5. Université de Montpellier, Centre d'Ecologie Fonctionnelle et Evolutive, France
  6. Université de Guyane, Laboratoire Écologie, Évolution, Interactions des Systèmes Amazoniens (LEEISA), France
9 figures, 4 tables and 1 additional file

Figures

Distribution of the sister-species M. helenor and M. achilles across South America.

The gray area represents the whole distribution of M. helenor, and the dotted area indicates the localities where M. helenor is in sympatry with M. achilles. Note that the different subspecies of M. helenor are found in different localities and display substantial variations in the proportion of black vs. blue areas on the dorsal sides of their wings. For example, M. h. bristowi and M. h. theodorus are allopatric in Ecuador, with their respective distribution separated by the Andes such that they are never in contact. While M. h. bristowi is found on the Pacific side of the country and has a wide blue band on the dorsal side of its wings, M. h. theodorus is found in Western Amazonia and displays a narrow blue band on the dorsal side of its wings. However, M. helenor is sympatric with M. achilles throughout the Amazonian rainforest. In French Guiana, the subspecies M. helenor helenor and M. achilles achilles display convergent dorsal color patterns with thin blue bands (Llaurens et al., 2021).

Figure 2 with 2 supplements
Differences in hue and brightness on the proximo-distal plane of Morpho wings.

(A) Illustration of the two protocols used to assess the differences in Morpho wing reflectance. The ‘Specular’ set-up allows for the quantification of the wing color variations, while the ‘tilt’ set-up can be used to quantify brightness variation for each Morpho wing (see Appendix 1 for extended methods). Variations of hue (B and C) and brightness (D and E) calculated from the wing reflectance measured with the ‘Specular’ set-up, and variations of brightness calculated from the wing reflectance measured from the ‘tilt’ set-up (F and G). Those hue and brightness parameters were calculated for the allopatric M. h. theodorus and M. h. bristowi (first column in orange and purple) and for the sympatric M. h. helenor and M. a. achilles (second column in green and blue) on the proximal-plane plane of their wings (I=illumination on the internal side of the wings, E=illumination on the external side of the wings). See Appendix 2—table 1 for the PERMANOVA analyses describing those graphs.

Figure 2—figure supplement 1
Differences in chroma on the proximo-distal and anteroposterior plane of Morpho wings.

Variation of the chroma parameter measured on the proximo-distal plane (A and B) and anteroposterior plane (C and D) calculated from the wing reflectance measured with “Specular” set-up. The chroma parameter was calculated for the sympatric M. h. helenor and M. a. achilles (first column in green and blue) and for the allopatric M. h. theodorus and M. h. bristowi (second column in orange and purple). The results of the permutation-based ANOVAs performed in order to test whether the sex, the taxa, or the angle of illumination has an effect on the estimated chroma are shown in (E). The chroma parameter is a proxy describing the intensity of the color reflected by the dorsal side of Morpho wings. As for every optical parameter measured, chroma is significantly different at every angle of illumination, suggesting that variations of chroma can be observed on the wings of Morpho butterflies (iridescence). Overall, we can see on every graph that the colors measured at a wide specular angle (30° to 45° in every tested direction) have the highest chroma, suggesting the existence of a more intense color signal at extreme angles of illumination. On the proximo-distal plane (A and B), chroma is significantly different between males and females. The graphs show that males tend to be more saturated than females, especially at extreme angles. On the anteroposterior plane, especially among the sympatric Morpho species analysis (C), the chroma of males is overall higher than the female’s, except at 45° angles of illumination where it drops for males but increases for females. The chroma parameter is always significantly different between M. h. bristowi and M. h. theodorus subspecies on both planes of illumination (B and D), and is also similar between sympatric M. h. helenor and M. a. achilles (A and C), consistent with convergence of the intensity of color saturation between sympatric species.

Figure 2—figure supplement 2
Differences in hue and brightness on the anteroposterior plane of Morpho wings.

Variations in hue (A and B) and brightness (C and D) calculated from the wing reflectance measured with the “specular” set-up on the anteroposterior plane, and variations of brightness calculated from the wing reflectance measured from the “tilt” set-up (E and F) on the anteroposterior plane. Differences in hue and brightness of the anteroposterior plane of the wings are shown for both the interspecific (first column in green and blue) and intraspecific comparisons (second column in orange and purple). The results of the permutation-based ANOVAs performed in order to test whether the sex, the taxa, or the angle of illumination has an effect on the estimated hue and brightness are shown in (G). In addition to the proximo-distal data presented in the main text (Figure 2), the analysis of the reflectance of Morpho wings on their anteroposterior plane shows that this plane is also iridescent as significant variations of hue and brightness are measured at different angles. The effect of sex on the variation of brightness is always significant no matter the pair tested or the method of measurement: the brightness of Morpho wings on this plane is thus sexually dimorphic (C, D, E, and F) as was found on the proximo-distal plane. However, the ‘tilt’ wing reflectance measurements of the brightness (E and F) are not as straightforward as the measurements taken on the proximo-distal plane showing that males were brighter than females (Figure 2F and G). Here, we observe that allopatric males are indeed brighter than females (F), but the Amazonian males from French Guiana are not as clearly different from their respective females (E). The difference in brightness between males and females in those two localities is also less important than the differences in brightness found between males and females measured on the proximo-distal plane (Figure 2F and G). This difference of brightness could be explained by the physical structures of the scales that could potentially better reflect light intensity on the proximo-distal plane than on the anteroposterior plane, generating more important shifts in brightness during a flapping flight motion. Finally, a significant effect of hue was found between the two allopatric populations of M. helenor (B) and between the two sympatric Morpho species (A). Conversely, no hue variation was found between the sympatric Morpho species on the proximo-distal plane (Figure 2B). Nevertheless, divergence in hue was found to be more important between the allopatric M. helenor subspecies than between the sympatric M. h. helenor and M. a. achilles, consistent with stronger convergence in coloration between sympatric species.

Figure 3 with 2 supplements
Characterization and perception of the iridescent coloration of Morpho butterflies.

(A) and (B): PCAs showing the variation in iridescence for both sexes, (A) in the two sympatric species from French Guiana (M. h. helenor vs. M. a. achilles) and (B) in the two allopatric Ecuadorian subspecies of M. helenor (M. h. theodorus vs. M. h. bristowi). Each point represents the global signal of iridescence of each individual, corresponding to the 21 complete reflectance spectra obtained from the 21 tested angles of illumination. The results of the PERMANOVA are shown on the top left corner of each graph. (C) The chromatic distances (i.e. the visual discrimination rate by a visual model) of the wing reflectance measured with the ‘Specular’ set-up on the proximo-distal plane. Visual modeling was used to calculate the chromatic contrast of blue coloration between allopatric M. helenor subspecies (red) and between the two sympatric sister-species M. helenor and M. achilles (green), as perceived by a Morpho visual system for every angle of illumination measured on the proximo-distal plane. The chromatic contrast likely perceived by UV-sensitive birds is shown in gray. Chromatic contrast of the female wings (top) and male wings (bottom). The threshold of discrimination is shown by the dotted line and set to 1 Just Noticeable Difference (JND). Error bars show the confidence intervals calculated during the bootstrap analysis.

Figure 3—figure supplement 1
Achromatic distances of the wing reflectance measured with the ‘Specular’ set-up on the anteroposterior and proximodistal plane.

Achromatic distances (i.e. the visual discrimination rate of brightness by a visual model) of the wing reflectance measured with the ‘Specular’ set-up on (A) the anteroposterior plane and (B) the proximodistal plane. We used visual modeling to calculate the achromatic contrast found between the blue coloration of the two allopatric populations of M. helenor sampled in Ecuador, as seen by a Morpho visual system (in red) and between the blue coloration of two species of Morphos from French Guiana, as seen by a Morpho visual system (in green). We also added the achromatic contrast of the Morpho wings as seen by the visual system of an avian predator (in gray). We separately measured the achromatic contrast of the female wings (first row of each figure) and male wings (second row of each figure) to account for sexual dimorphism. The dotted line represents the threshold of discrimination by any visual model: visual discrimination is considered possible if the measured chromatic distance is superior to the threshold. The overlap between the confidence intervals and the discrimination threshold shows that neither a bird visual model nor a Morpho visual model could discriminate between the wing brightness of Morphos from allopatric and sympatric populations.

Figure 3—figure supplement 2
Chromatic distances of the wing reflectance measured with the ‘Specular’ set-up on the anteroposterior plane.

We used visual modeling to calculate the chromatic contrast found between the blue coloration of the two allopatric populations of M. helenor sampled in Ecuador, as seen by a Morpho visual system (in red) and between the blue coloration of two species of Morphos from French Guiana, as seen by a Morpho visual system (in green). We also added the chromatic contrast of the Morpho wings as seen by the visual system of an avian predator (in gray). We separately measured the chromatic contrast of the female wings (first row) and male wings (second row) to account for sexual dimorphism. The dotted line represents the threshold of discrimination by any visual model: visual discrimination is considered possible if the measured chromatic distance is superior to the threshold. Overall, the results are similar to the chromatic distances measured on the reflectance spectra extracted from the proximo-distal plane of the Morpho wings (Figure 3C in the main text): the minute hue differences observed between species on the anteroposterior plane of the wings cannot be discriminated in a Morpho visual model, whereas the intraspecific divergent hues measured on the wings of allopatric M. h. bristowi and M. h. theodorus can be discriminated by a Morpho visual model.

Figure 4 with 2 supplements
Morpho male preference based on visual cues alone.

Probabilities of (A) approaching and (B) touching a con-subspecific female for M. h. bristowi males (purple) and M. h. theodorus males (orange) from Ecuador, as well as the probabilities of (C) approaching a con-specific female for M. h. helenor males (blue) and M. a. achilles males (green) from French Guiana, measured during experiment 4. The dotted line indicates the expected probability of approaching/touching a model if no preference is present. The p-values resulting from the binomial GLMM testing for preferentially directed interactions toward one female dummy are shown next to the corresponding graphs.

Figure 4—figure supplement 1
Morpho male pattern-based discrimination (experiment 2).

Probabilities of (A) approaching and (B) touching a con-subspecific female for M. h. bristowi males (purple). The M. h. bristowi males had the choice between a wild-type M. h. bristowi female model (on the right) and a modified M. h. bristowi female model with a narrowed blue band pattern like M. h. theodorus (on the left). The dotted line indicates the probability of approaching/touching a model expected if no preference was found. The p-values of Wilcoxon tests testing the significant departure from the 0.5 probability are shown next to the corresponding graphs.

Figure 4—figure supplement 2
Morpho male color-based discrimination (experiment 3).

Probabilities of (A) approaching and (B) touching a con-subspecific female for M. h. theodorus males (orange). The M. h. theodorus males had the choice between a WT M. h. theodorus female model and a modified M. h. bristowi female model with a narrow blue band pattern like M. h. theodorus. The dotted line indicates the probability of approaching/touching a model expected if visual preferences are lacking. The p-values of Wilcoxon tests testing the significant departure from the 0.5 probability are shown next to the corresponding graphs.

Figure 5 with 1 supplement
Divergent chemical compounds found on the genitalia of sympatric Morpho males.

nMDS representation of the differences in the concentration of (A) C8 to C16 chemical compounds and (B) C16 to C30 chemical compounds found in the genitalia of M. h. helenor and M. a. achilles males and females, calculated using Bray-Curtis distances. M. h. helenor are shown in blue (males are in dark blue and females in light blue), and M. a. achilles are shown in green (males are dark green and females are light green). The result of the PERMANOVA (999 permutations) testing the effect of sex and species on the chemical composition of M. h. helenor and M. a. achilles is shown on each figure.

Figure 5—figure supplement 1
Chromatograms ilustrating the chemical compounds found on male and female genitalia.

Example of a chromatogram obtained when comparing the chemical compounds from C16 to C30 found on male genitalia (A) and on female genitalia (B). The red chromatogram shows the chemical compounds of M. h. helenor individuals and the blue one the chemical compounds of the M. a. achilles individuals. Arrows point at specific compounds only found in M. h. helenor males (red arrows) or M. a. achilles males (blue arrows).

Experimental design of the four male mate choice experiments.

Experimental design of the male choice experiments performed between the Morphos originating from Ecuadorian populations of M. helenor (eastern population: M. h. theodorus and western population: M. h. bristowi, experiments 1–3) and from French Guiana (M. h. helenor and M. a. achilles, experiment 4), along with the associated preferences and visual cues tested. Note that the same males were used for experiments 1, 2, and 3.

Appendix 1—figure 1
Illustration of the set-up used to measure the reflectance of the wings of Morpho butterflies at different angles of illumination and observation.

We used a fiber holder to precisely control the angle between the two fibers. We ensured the measured surface was flat with a support holding the wings.

Appendix 1—figure 2
Scheme of the specular angles of illumination used to measure the reflectance of the wings.

(A) shows the coverage of the 13 angles of illumination measured for each wing. The angles analyzed on the proximo-distal plane are represented in (B), and the angles analyzed on the antero-posterior plane are shown in (C). Each letter used to describe the angles (I, E, P, A) refers to the side the light was directed to (Internal, External, Posterior, and Anterior side of wings, respectively).

Appendix 1—figure 3
Scheme of the angles measured in the tilt set-up.

The red dotted lines represent the specular 30° angle measured in the ‘specular’ set-up. The angular span between the two fibers is kept the same and the two fibers are tilted toward the normal of the wings (annotated with +) or away from it (annotated with -), allowing the additional measurement of the wing’s reflectance. This operation is repeated on the Internal, External, Anterior, and Posterior sides (I, E, A, and P, respectively) of the wings.

Tables

Table 1
Tetrad experiment results.

Number of mating events involving the different possible pairs of males and females from the two subspecies of M. helenor (bristowi vs. theodorus) in 30 tetrad experiments.

M. h. theodorus femaleM. h. bristowi female
M. h. theodorus male63
M. h. bristowi male516
Appendix 2—table 1
Permutation-based ANOVAs performed in order to test whether sex, taxa, or the angle of illumination have an effect on the estimated hue and brightness measured on the proximo-distal plane of Morpho wings.

We observe that the effect of sex is always significant on the variations of brightness hinting at sexual dimorphism. Hue is also significantly different between the allopatric sub-species of M. helenor, M. h. theodorus, and M. h. bristowi, whereas it was very similar for the two sympatric species M. h. helenor and M. a. achilles, suggesting convergence in wing coloration between the two sympatric species.

Hue Specular French GuianaHue Specular EcuadorBrightness Specular French GuianaBrightness Specular EcuadorBrightness 30° tilt French GuianaBrightness 30° tilt Ecuador
Sex0.0620.001 ***0.002 **0.001 ***0.001 ***0.001 ***
Species0.1880.001 ***0.001 ***0.0660.5060.007 **
Angle0.001 ***0.001 ***0.001 ***0.001 ***0.001 ***0.001 ***
Sex:Species0.030 *0.002 **0.001 ***0.001 ***0.020 *0.114
Species:Angle0.4850.6380.1880.3620.006 **0.050 *
Appendix 3—table 1
Annotation of the chemical compounds significantly associated with both sexes of M. helenor or M. achilles using an Indicator value analysis.

We used an Indicator Value analysis to find the chemical compounds allowing us to discriminate the males and females of M. helenor and M. achilles separately. Because we used a protocol allowing for the detection of small peaks during the MZmine analysis to not exclude ‘pheromone-like’ molecules, our analysis is sensitive to small spectral variations. This could explain why some compounds were associated with different (but similar) MZmine-detected occurrences. X refers to an undetermined double bond position or configuration (Z or E). The ‘Ions’ column describes the spectrum with the major ion, followed by other ions in intensity order, and the underlined visible ion corresponding to the molecular ion.

SexDatasetAnnotationLRIIonsSpeciesp-value
MalesC8 to C16(E)-β-Ocimene104893; 79; 41; 136M. helenor0.0491
Phenyl ethyl alcohol111291; 122; 65M. achilles0.0004
Unknown compound 1112380; 52; 73; 98; 124M. helenor0.0311
Ethyl octanoate119588; 57; 101; 43; 73; 127; 172M. helenor0.0001
Tetradec-1-ene139043; 55; 69; 83; 97; 196M. achilles0.0282
Ethyl decanoate139488; 101; 43; 73; 155; 200M. helenor0.0001
Dodec-x-enol145455; 68; 41; 82; 96; 184M. helenor0.0002
Ethyl dodecanoate159388; 101; 43; 73; 155; 228M. helenor0.0001
(x)-Ethyl Tetradec-x-enoate176488; 55; 41; 96; 166; 254M. helenor0.0001
(x)-Tetradec-x-enyl acetate178443; 68; 54; 82; 96; 194M. helenor0.0001
C16 to C30Ethyl hexadecanoate199388; 101; 43; 73; 155; 284M. helenor0.0001
Unknown compound 2204644; 207; 49; 83; 55; 69M. helenor0.0001
Geranyl decanoate214869; 93; 41; 121; 136; 308M. helenor0.0001
(Z)-Ethyl Octadec-9-enoate216855; 41; 69; 88; 83; 310M. helenor0.0001
Ethyl octadecanoate219388; 101; 43; 55; 157; 312M. helenor0.0001
Unknown compound 3310944; 207; 57; 71; 85; 281M. achilles0.0096
(x)-Tetradec-x-enyl hexadecanoate314868; 82; 96; 194; 43; 450M. helenor0.0001
Unknown compound 43342207; 43; 55; 73; 81; 95M. achilles0.0159
(x)-Tetradec-x-enyl octadecanoate335068; 82; 96; 194; 43; 57M. helenor0.0001
FemalesC8 to C16Dec-1-ene99056; 41; 70; 83; 97; 140M. helenor0.028
Dodec-1-ene119043; 55; 69; 83; 97; 168M. helenor0.0302
Appendix 4—table 1
Repeatability (R) of the measurements of iridescence performed on the wings of M. h. bristowi, M. h. theodorus, M. h. helenor, and M. a. achilles.

The repeatability was measured for the values of Brightness, Hue, and Chroma at every angle of illumination.

RSEEmp.2.5%Emp.97.5%p-valueVariableAngle_ID
0.931953870.013073070.903335170.952364424.38E-77Chroma0
0.88159250.021946440.828073120.916427392.02E-58Brightness0
0.937314460.012643650.906491660.955523687.07E-80Hue0
0.814782310.032285710.739222020.866254591.06E-43Chroma15 A
0.670461720.052655380.548440620.755445741.20E-25Brightness15 A
0.910218410.016869650.870678910.934798521.06E-67Hue15 A
0.775599760.038068480.688136180.835310321.52E-37Chroma15E
0.67089990.050141850.562491930.749612191.10E-25Brightness15E
0.871019430.024902250.811479630.906789581.41E-55Hue15E
0.831433940.029379520.768033560.877255329.15E-47Chroma15I
0.735129520.040708020.639074040.80052032.51E-32Brightness15I
0.918395260.015789240.880396180.944337666.33E-71Hue15I
0.799820880.03502370.71623140.854561683.39E-41Chroma15 P
0.71989580.048678780.616513390.795266351.36E-30Brightness15 P
0.889304220.020205340.84434170.923314421.14E-60Hue15 P
0.794215990.035981340.71917630.85512822.62E-40Chroma30 A
0.677329380.049190370.567517560.760169642.80E-26Brightness30 A
0.92891880.013055970.899742570.951623241.33E-75Hue30 A
0.704445850.046680470.606698010.787771676.07E-29Chroma30E
0.711781830.0454880.613938970.791545881.03E-29Brightness30E
0.878746350.023311220.822966620.914098941.25E-57Hue30E
0.65509050.052642390.538160180.745077052.76E-24Chroma30I
0.745540830.042619230.645664560.813911481.41E-33Brightness30I
0.875820590.023477920.822280530.914843647.75E-57Hue30I
0.90418860.018785980.857739670.933606671.64E-65Chroma30 P
0.812050690.032800430.743028440.869914713.15E-43Brightness30 P
0.971124530.005850860.957464470.97957591.88E-106Hue30 P
0.854006730.02605580.797578030.897379061.77E-51Chroma45 A
0.770398820.03740880.690026710.835701778.13E-37Brightness45 A
0.902035760.019361480.855241260.930763249.14E-65Hue45 A
0.759131340.037865690.679162310.823516542.66E-35Chroma45E
0.817576290.028961740.754310070.870157133.40E-44Brightness45E
0.820712060.032003760.748737360.868272579.31E-45Hue45E
0.742591910.041709080.648650710.808939113.23E-33Chroma45I
0.711828550.044785490.622814110.78502761.02E-29Brightness45I
0.841042410.028868780.779250180.889416371.10E-48Hue45I
0.833457450.030610640.762214610.883167793.69E-47Chroma45 P
0.725762450.04351710.622980390.792671063.01E-31Brightness45 P
0.795372050.033488550.721964610.850995381.73E-40Hue45 P
0.760962230.038759620.674665420.826468121.53E-35Chroma30A+
0.570077860.059363610.440730310.680321816.32E-18Brightness30A+
0.829013030.030128310.758975090.875668132.67E-46Hue30A+
0.898303540.018593550.855664020.92603421.65E-63Chroma30E+
0.605906470.055314830.486159640.69227342.16E-20Brightness30E+
0.860408690.026523430.797822120.902848595.85E-53Hue30E+
0.831518770.030010950.767022060.879827238.82E-47Chroma30I+
0.802647730.033291810.726078190.856827161.18E-41Brightness30I+
0.757405080.039758580.672918850.817656254.46E-35Hue30I+
0.831499210.031233070.760717490.884202998.89E-47Chroma30P+
0.73156420.04184150.642050630.804113596.54E-32Brightness30P+
0.541805070.063473570.397732730.642647213.63E-16Hue30P+
0.825649330.031410510.757852060.878202621.15E-45Chroma30A-
0.703043310.045657090.6097970.781199948.47E-29Brightness30A-
0.704307950.047439610.597258790.779097076.27E-29Hue30A-
0.769534390.040632470.683002520.838112161.07E-36Chroma30E-
0.848007580.028192330.781569110.889728663.73E-50Brightness30E-
0.845279770.02681520.78548960.888803491.43E-49Hue30E-
0.91424090.015665380.878656150.938196663.01E-69Chroma30I-
0.832974970.028025450.769985440.880578364.59E-47Brightness30I-
0.887928460.021117320.840247280.920214212.95E-60Hue30I-
0.75962080.041140260.66985370.8267022.29E-35Chroma30P-
0.71154460.045681590.612029120.783549851.09E-29Brightness30P-
0.94125190.010895930.916578770.957029334.35E-82Hue30P-

Additional files

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Joséphine Ledamoisel
  2. Bruno Buatois
  3. Rémi Mauxion
  4. Christine Andraud
  5. Melanie McClure
  6. Vincent Debat
  7. Violaine Llaurens
(2025)
Convergent iridescence and divergent chemical signals in sympatric sister-species of Amazonian butterflies
eLife 14:RP106098.
https://doi.org/10.7554/eLife.106098.3