Color is a multifaceted trait with many functions such as predator defense, thermoregulation, and immune response. We investigated pupal color variation in Chlosyne lacinia pupae, focusing on identifying the cue for increased melanization. Through laboratory experiments, we demonstrated pupae reared on black backgrounds exhibited significantly higher melanization compared to those on white backgrounds. Additionally, black pupae experienced longer developmental periods, suggesting a trade-off between defense and developmental time. Our findings support crypsis as a likely evolutionary driver for increased melanization in response to substrate color. We discuss potential implications for predator avoidance, immune response, and developmental costs associated with melanization. This study provides insights into the adaptive significance of pupal melanization in response to environmental cues, shedding light on the complex interplay between life history traits in butterflies.

Identification of Black Pigment

Using the protocol for cuticular pigment extraction in Debecker et al. (2015), we used pupal exuviae from colony individuals to verify the black pigment as melanin. The pigment was extracted by adding 1 M NaOH/10% DMSO, of which the volume was adjusted to the mass of each exuvia (extraction volume in μL = 200 × exuvial mass in mg) after the exuviae had been homogenized twice for seven seconds. The samples were incubated at 80°C for two hours in a water bath and then centrifuged at 12,000 g for 10 minutes. The absorbance of the samples was read at 380 nm and compared to standard absorbance curves for melanin.

Cue Identification

Colony larvae of the third/fourth instar were selected for use in multiple experiments to identify the cue for pupal melanization. First, we conducted a shading experiment. We placed larvae in 32oz Styrofoam cups with white or black mesh over the opening held in place by the lid with a hole cut in the middle. A moist paper towel was placed at the bottom of each cup to maintain moisture levels and facilitate cleaning. Larvae were kept at room temperature (~23°C) under grow lights with an 18hrL:6hrD light cycle to mimic summer months. Larvae were fed ad libitum with A. trifida leaves with their cups checked and cleaned daily. There were 60 individuals with 30 in each treatment. Next, we tested the effects of larval density and temperature on pupal melanization. Using the same 32 oz Styrofoam cups, we tested 1 larva (low density), 3 larvae (medium density), and 6 larvae (high density) at each of two temperatures: 23°C and 30°C for a total of 60 individuals in this trial. 23°C were kept under grow lights and 30°C were housed in a Thermo Scientific P/N 3185051 Rev. 2 incubator at 30°C, ~30% humidity; both at a 18hrL:6hrD light cycle. Common sunflowers were placed in each cup as food and were checked daily. We also tested the effects of wounding on pupal cuticular melanization. The same habitat conditions as the shading experiment were used, but for 20 larvae. We used a sterilized sewing needle to wound 10 of the 20 individuals and then documented the pupal coloration. All larvae from these trials (shading, temperature and density, and wounding) categorically resulted in 100% light-colored pupae, so we did not conduct any further experiments past the pilot trials (Supplemental Table 1).

Additionally, we tested the effect of substrate color on pupal coloration. Twenty larvae were selected at random and placed in 32oz Styrofoam cups painted inside with either white or black non-toxic paint. A moist paper towel was placed in the bottom of each cup to keep the treatment environment moist and facilitate cleaning. A hole was cut into the plastic lid and mesh of the corresponding cup color was placed over the cup opening before the lid was placed on that cup covering almost the entire exposed surface area to the larva with the specified treatment color. The same methodology was used as the 30°C cups from the larval density and temperature experiment except larvae were fed with A. trifida leaves. Upon pupation, the date and color of the pupa were noted. On the second day of the pupal stage, a photograph was taken with an Olympus R-C-OLY-IM015 digital camera using a standardized photography setup (Stevens et al., 2007). The day of eclosion was noted and each individual was sexed. The trial was repeated a second time to verify the results.

The photographs were uploaded to ImageJ (ImageJ.JS v 0.4.0) and converted to a 2-D binary (black and white) image. Given the pupal coloration per individual is symmetrical, a lateral orientation of the left dorsal view of the pupa was used for the photographs (Fig. 1). Within the binary image, we obtained the information for the total number of pixels per pupa, the number of white pixels, and the number of black pixels. This information was used to create a percentage of melanization for each pupa.

Data Analysis

Using R (version 4.3.2) (R Core Team, 2024), we conducted Generalized linear models (GLM) to test the effect of the substrate color (Treatment) and sex, and their interactions, on pupal development time, and percentage of melanization. Based on the output of these statistical tests, we conducted an AIC using the AICcmodavg package (Mazerolle, 2023) to select the model of best fit that explains the percent of melanization in pupae. We also conducted ANOVAs and GLMs to test if the percentage of melanization in pupae, sex, and larval color affected the pupal development time.

Additionally, we wanted to test if there were effects of sex, percentage of melanization, and larval color on pupal developmental time. We used the dplyr package (Wickham et al., 2023) and grep function to subset the data for both white and black pupae. We then conducted GLMs to test these effects. Furthermore, we conducted a generalized linear regression analysis to test for the effect of pupal melanization on pupal developmental time.