Wing-pattern-specific effects of experience on mating behavior in Heliconius melpomene butterflies

TL;DR: The results suggest that social experience can influence male mating behavior in H. melpomene and that behavioral plasticity may differ across populations in this species.

Abstract: Many animals have the ability to learn, and some taxa have shown learned mate preference. This learning may be important for speciation in some species. The butterfly Heliconius melpomene is a model system for several areas of research, including hybridization, mate selection, and speciation, partially due to its widespread diversity of wing patterns. It remains unclear whether these butterflies can learn to prefer certain mates and if social experience shapes realized mating preferences. Here we test whether previous experience with a female influences male mate preference for two different H. melpomene subspecies, H. m. malleti and H. m. rosina. We conducted no-choice behavioral assays to determine if latency to court and whether males courted (vs no courtship) differed between naive males and males with previous exposure to a young, sexually mature, virgin female. To test whether assortative courtship preference is learned in H. melpomene, males were either paired with a female who shared their phenotype or one who did not. Naive H. m. malletti males courted assortatively, while naive H. m. rosina males did not. Experienced H. m. malleti males reduced their courting relative to naive males, suggesting that social experience with a sexually mature female that does not result in copulation may be perceived as a negative experience. In contrast, experienced H. m. rosina males exhibited similar courting rates to naive H. m. rosina males. Our results suggest that social experience can influence male mating behavior in H. melpomene and that behavioral plasticity may differ across populations in this species.

Summary

Introduction

• Many of the behaviors and decisions that an animal makes are affected by its observations and capacity to learn.
• Learning in animals is often complex and is likely the result of the social dynamics and settings of a species (Coussi-Korbel and Fragaszy, 1995) .
• Previous studies show that mimetic coloration in this species is important in choosing mates, and that these butterflies show assortative mating when choosing between their own and a different, closely related species (Heliconius cydno) (Jiggins et al., 2001) .
• 3) If males are not able to learn, then courting was predicted to occur at random in both experienced and naïve males.

Study species and husbandry

• Heliconius melpomene is a widespread neotropical butterfly found in Central and South America (Brower, 1994; Sheppard et al., 1985) .
• The species is well known for its high diversity in color patterns, which play an important role in speciation (Jiggins et al., 2004) .
• Caterpillars from the colony were given Passiflora plants ad libitum, and prior to pupation, plants containing caterpillars were removed from the breeding cages and moved to a separate 60.96 x 60.96 x 142.24 cm cage until butterfly emergence from pupa.
• The greenhouse was lit by Sun Blaze T5 high output 120-volt fluorescent light fixtures (containing UV wavelengths), in addition to natural sunlight, and the presence of UV light in the greenhouse was confirmed using an Ocean Optics Jaz spectrometer.
• Males were placed into sex-and phenotype-specific cages, so they were isolated from both females and other wing patterns prior to behavioral assays.

Observational Experiment Time of Day Selection

• To determine the time of day when the butterflies were the most active, the authors observed butterflies in colony cages for three consecutive days, between 6:00 am and 8:00 pm.
• Point counts were conducted every thirty minutes, where behaviors (flight, walk, flutter, abdomen lift, bask [defined by resting with wings held in open position], rest [defined by resting with wings held in closed position], antennae wiggle, court, and copulate) were recorded for each cage, followed by two ten-minute focal watches of one male and one female butterfly selected at random.
• Based on observations, the authors determined that butterflies were most active between the hours of 10:00 am to 2:00 pm.

Behavioral Watches

• All behavioral watches took place between 10:00 am and 2:00 pm, the time of peak H. melpomene activity in their greenhouse.
• Each watch consisted of a male aged ten or twelve days old, and a female between three and five days old.
• To test whether males courted females with matching wing patterns faster than they courted conspecific females with dissimilar wing patterns, the authors tested latency to courtship and presence of courtship of naïve, 12-day-old H.m. malleti and H.m. rosina males matched with either females of their own phenotype or females of different phenotypes.
• Afterward, the female was removed, and the male was returned to the all-male, phenotype-specific cage.
• The number of incidents of each type of behavior (flight, walk, flutter, abdomen lift, bask (wings open), rest (wings closed), antenna wiggle, sitting near, and court) were recorded.

Statistical Analyses

• All statistical analyses were performed in JMP v. 14 (SAS Institute, Cary, NC, U.S.A.).
• The authors assessed whether latency to court was influenced by male experience or female wing pattern (similar or different from the male's) using a GLM with male experience and female wing pattern as factors, as well as an interaction term.
• To test whether female behavior during a male's first experience with a female had an effect on the observed courtship behavior in later interactions with females, the authors analyzed all behavioral data collected on day 10 watches (N=51 watches with behavioral data) and examined whether any of these behaviors were predictive of male courting on day 12.
• To do this the authors ran a principal components analysis on all the female behaviors and then ran logistic regression models on the first three principal components.

Ethical Note

• All H. melpomene butterflies were kept under laboratory conditions as defined by U.S. Department of Agriculture, Animal and Plant Health Inspection Service permit P526P-17-00343.
• Before being used in behavioral watches all butterflies were maintained in cages in a climatecontrolled setting in conditions similar to those of their native habitat, and cages were inspected daily for ample food and appropriate conditions.

Discussion

• The authors results show that male H. melpomene butterflies change their mating behavior in response to a social experience.
• The negative effect of the pre-mating social exposure, and the wing-pattern-specific response to this pre-mating social exposure, were unexpected.
• These males are then tested repeatedly, and past experience is often not accounted for when male preference is assessed, assuming that past experience does not inform present courting decisions.
• It also highlights the importance of checking for both positive and negative valence when testing the presence of learning.

Conclusion

• Here the authors show that male H. melpomene butterflies use past social experience to inform current mating behavior.
• This response is lineage (wing pattern) specific, and coincides with lineagespecific differences in male assortative preference.
• The authors findings strongly suggest that there are lineage-specific selective forces acting on cognitive function in Heliconius butterflies.
• Lineage-specific effect of experience on male courtship.

Figures

Table 2. Female behavior during early exposure did not influence male likelihood of courting in later female encounters. Test statistics and p-values from logistic regression models using composite behavioral variables PC1, PC2, PC3. N=51.

Table 1. Loadings for principle components from PCA for female behavior during the training period for day 10 males.

Full text

1
Wing-pattern-specific effects of experience on mating behavior in Heliconius
1
melpomene butterflies
2
Peyton A. Rather*, Abigail E. Herzog*, David A. Ernst, Erica L. Westerman
3
Department of Biological Sciences, University of Arkansas, Fayetteville, AR 72701
4
*
equal contributors
5
6
7
Abstract
8
Many animals have the ability to learn, and some taxa have shown learned mate preference. This
9
learning may be important for speciation in some species. The butterfly Heliconius melpomene is
10
a model system for several areas of research, including hybridization, mate selection, and
11
speciation, partially due to its widespread diversity of wing patterns. It remains unclear whether
12
these butterflies can learn to prefer certain mates and if social experience shapes realized mating
13
preferences. Here we test whether previous experience with a female influences male mate
14
preference for two different H. melpomene subspecies, H. m. malleti and H. m. rosina. We
15
conducted no-choice behavioral assays to determine if latency to court and whether males
16
courted (vs no courtship) differed between naïve males and males with previous exposure to a
17
young, sexually mature, virgin female. To test whether assortative courtship preference is
18
learned in H. melpomene, males were either paired with a female who shared their phenotype or
19
one who did not. Naïve H. m. malletti males courted assortatively, while naïve H.m. rosina males
20
did not. Experienced H. m. malleti males reduced their courting relative to naïve males,
21
suggesting that social experience with a sexually mature female that does not result in copulation
22
may be perceived as a negative experience. In contrast, experienced H. m. rosina males exhibited
23
.CC-BY-NC-ND 4.0 International licenseavailable under a
was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted July 17, 2020. ; https://doi.org/10.1101/2020.07.15.205435doi: bioRxiv preprint

2
similar courting rates to naïve H. m. rosina males. Our results suggest that social experience can
24
influence male mating behavior in H. melpomene and that behavioral plasticity may differ across
25
populations in this species.
26
27
Keywords
28
behavioral plasticity, mate choice, Lepidoptera, assortative mating, social learning, male choice
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
.CC-BY-NC-ND 4.0 International licenseavailable under a
was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted July 17, 2020. ; https://doi.org/10.1101/2020.07.15.205435doi: bioRxiv preprint

3
Introduction
47
48
Many of the behaviors and decisions that an animal makes are affected by its observations and
49
capacity to learn. Learning can be defined as a set of processes that allows an animal to acquire,
50
store, and use information gathered from the environment (Galef and Laland, 2005). Learning in
51
animals is often complex and is likely the result of the social dynamics and settings of a species
52
(Coussi-Korbel and Fragaszy, 1995). There is a substantial amount of evidence that animals have
53
the ability to socially learn (Dukas, 1998). Some of the many behaviors that might be the result
54
of social learning include food choices, predator avoidance, and mate preferences. For example,
55
many species of fish have been observed to learn how to find food, how to recognize predators,
56
and how to assess mate quality (Brown and Laland, 2003). This breadth of learning ability,
57
however, is not limited to vertebrates (Dukas, 2008; Verzijden et al., 2012).
58
59
It is now understood that learning affects many essential activities of invertebrates, including
60
predator avoidance and social interactions (Dukas, 2008, 2010). Particularly, many insects and
61
spiders have shown the ability to learn mate preference. Studies on the wolf spider Schizocosa
62
uetzi have shown that female social experience in their penultimate juvenile period can affect
63
their mate choices as adults (Hebets, 2003). Female Teleogryllus oceanicus crickets modify
64
their mate preferences after hearing attractive male songs (Bailey and Zuk, 2009), and female
65
Bicyclus anynana butterflies learn preferences for enhanced male ornaments (Westerman et al.,
66
2012). Male B. anynana also learn preferences for wing pattern elements in females (Westerman
67
et al., 2014). Furthermore, work with Drosophila melanogaster fruit flies have shown that
68
learning to be selective leads to a higher lifetime mating success than males who court
69
.CC-BY-NC-ND 4.0 International licenseavailable under a
was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted July 17, 2020. ; https://doi.org/10.1101/2020.07.15.205435doi: bioRxiv preprint

4
indiscriminately (Dukas et al., 2006). Therefore, when it comes to mate preference and sexual
70
behavior in insects it is often beneficial to learn.
71
72
Learning can potentially increase rates of assortative mating, which can lead to speciation
73
through processes such as when young animals imprint on parents (Dukas, 2013). One such
74
example of this is how cross
‐
fostering experiments in two subspecies of zebra finch
75
demonstrated that assortative mating is due to imprinting. Birds in this study paired with mates
76
that resembled their foster parents instead of their own phenotype (Irwin and Price, 1999). It has
77
also been shown that mate preference can be learned in mature animals, such as male guppies
78
and Syrian hamsters. These animals have demonstrated learning to discriminate against
79
heterospecific mates after courtship interactions (Verzijden et al., 2012). This type of learning
80
would help maintain speciation. With these studies in mind, we might expect that Heliconius
81
butterflies, or other animals with high levels of speciation, might learn to court assortatively.
82
83
Heliconius butterflies have a long lifespan compared to other species of butterflies, which allows
84
them to potentially mate multiple times (Gilbert, 1972). Therefore, the ability to learn in
85
response to mating experiences could be advantageous. Studies have shown that male mate
86
preferences evolve early in the speciation process in Heliconius within both intraspecific hybrid
87
mating zones and conspecific polymorphic populations (Merrill et al., 2011a). These male mate
88
preferences are based on wing color pattern cues, which are under natural selection to correspond
89
to local mimetic environments (Gray and McKinnon, 2007; Kronforst et al., 2006). Heliconius is
90
well known for its diversity in color patterns, and divergence in these color morphs is associated
91
with speciation and adaptive radiation (Heliconius-Genome-Consortium* et al., 2012).
92
.CC-BY-NC-ND 4.0 International licenseavailable under a
was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted July 17, 2020. ; https://doi.org/10.1101/2020.07.15.205435doi: bioRxiv preprint

5
93
Here we take advantage of the social butterfly species Heliconius melpomene, whose widespread
94
diversity of color patterns makes it an ideal model for studies on speciation and mating patterns
95
(Jiggins et al., 2004). In this species, mimetic color patterns play a key role in species
96
recognition, and mate preferences based on these patterns evolve alongside changes in wing
97
pattern (Jiggins et al., 2004). Previous studies show that mimetic coloration in this species is
98
important in choosing mates, and that these butterflies show assortative mating when choosing
99
between their own and a different, closely related species (Heliconius cydno) (Jiggins et al.,
100
2001). Furthermore, males often do discriminate between conspecific females with different
101
wing patterns, and do not copy the mate preferences of conspecific males who have different
102
wing patterns (Jiggins et al., 2004). However, it remains unclear whether individual H.
103
melpomene males use past social experience with sexually receptive (or non-receptive) females
104
to inform current mating decisions. The ability to learn mate preferences for intraspecies
105
variation in wing pattern may be important for the initiation of assortative mating, reproductive
106
isolation, and the speciation process.
107
108
Here we test whether experience impacts future male mate preference and courting behavior in
109
two races of H. melpomene using three distinct H. melpomene color morph phenotypes (Figure
110
1). We had three alternative hypotheses: 1) If males learn, then we predicted that experienced
111
males would be more likely to court and have a shorter latency to court relative to naïve males.
112
This type of learning is seen in B. anynana, where males exposed to dorsal hindwing spot
113
number variation learn preferences for this trait (Westerman et al., 2014). 2) If however male
114
exposure to a female is somehow a negative experience, then we predicted that experienced
115
.CC-BY-NC-ND 4.0 International licenseavailable under a
was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted July 17, 2020. ; https://doi.org/10.1101/2020.07.15.205435doi: bioRxiv preprint