Datasets and Rcode for the manuscript: "Which fruit traits correlate reliably with sugar content, and does this vary with dispersal syndrome?"

Deciding which fruits to eat, and when, are key decisions made by frugivores, with implications for animal nutrition, seed dispersal and other ecosystem interactions. It has long been hypothesized that fruit traits, including colour, odour and hardness, are adapted to attract frugivores and facilitate fruit foraging by acting as honest signals of ripeness and nutritional reward. These have been shaped by selective pressures from different guilds of frugivores according to their sensory systems, yet, how traits change during ripening in fruits with different putative dispersal syndromes is poorly understood. We predicted visual traits to be more correlated with changes in sugar content during ripening in a bird-consumed fruit, and odour and hardness traits to be correlated with sugar content in mammal-consumed fruits. We measured colour via image analysis, odour via solid-phase microextraction GC-MS, hardness via a portable mechanical tester, and sugar content using a digital refractometer in n = 100 bananas (Musa acuminata) and n = 160 mangoes (Mangifera indica), two large, climacteric fruits as models of mammal-dispersed fruits, and in n = 80 black currants (Ribes nigrum), a small, non-climacteric berry as model for bird-dispersed fruit. Bananas, and to a lesser extent, mangoes, tend to get softer, have higher red-green and blue-yellow chromaticity values and become more odiferous as they get sweeter. Currants get softer, have higher red-green and blue-yellow chromaticity values, lower luminance, and become less odiferous as they become sweeter. Our results demonstrate similarities and variation in fruit trait correlations among plant species that suggest possible differences between dispersal syndromes, as well as more universal traits. We provide new methods and data on the interrelatedness of fruit traits, and contribute to the understanding of fruiting plant signal evolution and plant-animal interactions. Future studies could usefully build on these insights by incorporating a large and phylogenetically-controlled dataset.