Fruit colour conceals endocarp dimorphism from avian seed dispersers in a tropical beach plant, Scaevola taccada (Goodeniaceae), found in Okinawa

• Published in: Journal of tropical ecology Vol. 31; no. 4; pp. 335 - 344
• Main Authors: Tanaka, Keita D., Denda, Tetsuo, Ueda, Keisuke, Emura, Naoko
• Format: Journal Article
• Language: English
• Published: Cambridge, UK Cambridge University Press 01.07.2015
• Subjects: Birds; Dispersal; Fruits; Goodeniaceae; Leaves; Nutritive value; Plant populations; Predation; Scaevola taccada; Seed dispersal; Seeds; Statistical methods; avian seed dispersal; deception; avian visual model; interspecific communication; cryptic dimorphism
• ISSN: 0266-4674; 1469-7831
• DOI: 10.1017/S0266467415000218

Theory predicts that honest signalling strategies will not always be evolutionarily stable in interspecific communication, yet to demonstrate such a transition of signalling modality between honesty and dishonesty in the wild would be difficult. An endocarp dimorphism has been found in Scaevola taccada fruits: a morph with a cork substrate that facilitates ocean current seed dispersal and a morph without the cork. Both types of fruit are covered with sugar-containing flesh, and are similar in size and colour to one another (at least from a human perspective). The apparent lack of external differences between morphotypes could potentially degrade mutualistic relations between the plant and seed-dispersing birds because the presence of a cork could lower the fruit's nutritional value. Thus, unless seed dispersers can discriminate between the different types of fruit, this system may provide an example of a transition between honest and dishonest signalling. We examined S. taccada fruit and leaf colours from an avian visual perspective. Even though the fruits and leaves were different in colour from one another to birds, there was no perceivable difference in the colours between fruit morphotypes. Therefore, fruit colour is not an honest indicator of reward to seed dispersers. Further, we propose an adoption of a statistical method in avian visual modelling studies that avoids the common statistical errors, such as violation of the congruence principle.