Camouflage in lichen moths: Field predation experiments and avian vision modelling demonstrate the importance of wing pattern elements and background for survival

Abstract

Background matching is perhaps the most ubiquitous form of defensive camouflage in the animal kingdom, an adaptive strategy that relies on the visual resemblance between a prey organism and its background to promote concealment from predators. The importance of background matching has been acknowledged for over a century, yet despite its renown and apparent pervasiveness, few studies exist that have objectively quantified its occurrence and tested the functional significance of background matching in a specific animal study system. The North Island lichen moth Declana atronivea presents a fascinating system to investigate such anti-predator coloration. This species possesses high contrast black and white forewings that appear to resemble lichen. Here we assessed the contribution of background matching to the antipredator defence of D. atronivea using field predation experiments with realistic models. We found that D. atronivea coloration confers a significant survival advantage against native avian predators when on lichen backgrounds compared to bark backgrounds, with an intermediate level of predation occurring when models were near, but not on lichen. This suggests that D. atronivea wing patterns are an adaptation for background matching. We subsequently used calibrated digital photography, avian vision modelling and image analysis techniques to objectively quantify the degree of background matching exhibited by D. atronivea and assessed the contribution of different visual elements (colour, luminance and pattern) to camouflage in this species. Only the pattern elements of D. atronivea presented a close match to that of the lichen backgrounds, with both chromatic and achromatic cues found to be poor predictors of background matching in this species. This study is one of the first to integrate vision modelling, quantitative image analysis and field predation experiments using realistic models to objectively quantify the level and functional significance of background matching in a real species, and presents an ideal system for further investigating the interrelation between multiple mechanisms of camouflage.

Keywords: artificial models; background matching; camouflage; colour analysis; crypsis; lichen resemblance; moth.

FIGURE 1

Declana atronivea moths have intricate black and white forewing patterns that appear to resemble lichen and may have an adaptive function for camouflage through background matching.

FIGURE 2

Placement of the artificial moth models for the (a) lichen treatment, (b) off‐lichen treatment and (c) bark treatment.

FIGURE 3

Survival curves of moth models on the bark (solid purple line), lichen (dashed orange line) and off‐lichen (dashed blue line) backgrounds. The curves represent the probability of surviving predation by avian predators as a function of time and are based on Kaplan–Meier survival estimates that account for censoring due to survival to the end of the study. Overall, models on bark had significantly lower survival (0.48 ± 0.05) compared to those on lichen (0.65 ± 0.04) (p = 0.01) at the end of the 72‐h period, whereas the survival rate of off‐lichen (0.55 ± 0.05) did not differ significantly from bark (p = 0.23) nor lichen (p = 0.18).

FIGURE 4

Chromatic (colour) discrimination values (JNDs) for moths on the different background substrates as modelled under blue tit vision. Moths placed on kowhai had the lowest JND values (mean JND = 2.09 ± 0.09), suggesting that they would be difficult to distinguish from the background under the current avian vision model. All other substrates presented poor colour matches to the moths, particularly totara (mean JND = 7.44 ± 0.25) and lichen‐kauri (mean JND = 6.79 ± 0.53). Boxplots show median values (middle line), interquartile range (box) and the range values including some outliers (dots which extend beyond the min and max of the boxplot). The horizontal grey dashed line indicates JND = 3 for comparison.

FIGURE 5

Achromatic (luminance) discrimination values (JNDs) for moths on the different background substrates as modelled under blue tit vision. All moth and background combinations presented apparently poor achromatic matches (all above 3 JND), suggesting that Declana atronivea would likely be distinguishable from the substrates under the current avian vision model. Moths on lichen‐kauri and lichen‐titoki had the lowest JND values (mean JND = 6.31 ± 0.77 and 6.99 ± 0.47 respectively), and titoki presented the highest JND values (mean JND = 28.3 ± 0.60). Boxplots show median values (middle line), interquartile range (box) and the range values including some outliers (dots which extend beyond the min and max of the boxplot). The horizontal grey dashed line indicates JND = 3 for comparison.

FIGURE 6

Pattern energy difference (PED) values for moths on the different background substrates. Lower values indicate a closer match and thus better background pattern matching. Here, moths had the lowest PED values on the two lichen substrates (both PED ± SE = 0.033 ± 0.003), while all of the bark substrates presented significantly higher differences in pattern energy, suggesting much better background pattern matching on lichen compared to bark backgrounds. Boxplots show median values (middle line), interquartile range (box) and the range values including some outliers (dots which extend beyond the min and max of the boxplot).