1335 Fruit color and contrast in seasonal habitats – a case study from a cerrado savanna Maria Gabriela G. Camargo, Eliana Cazetta, H. Martin Schaefer and L. Patrícia C. Morellato M. G. G. Camargo (gabicamargo@yahoo.com) and L. P. C. Morellato, Depto de Botânica, Laboratório de Fenologia, Grupo de Fenologia e Dispersão de Sementes, Inst. de Biociências, Univ. Estadual Paulista (UNESP), Avenida 24A 1515, CEP 13506-900, Rio Claro, SP, Brazil. – E. Cazetta, Depto de Ciências Biológicas, Universidade Estadual de Santa Cruz, Rodovia Ilhéus-Itabuna km 16, CEP 45662-900, Ilhéus, BA, Brazil. – H. M. Schaefer, Dept of Evolutionary Biology and Animal Ecology, Faculty of Biology, Univ. of Freiburg, Hauptstr. 1, DE-79104 Freiburg, Germany. Communication contributes to mediate the interactions between plants and the animals that disperse their genes. As yet, seasonal patterns in plant–animal communication are unknown, even though many habitats display pronounced seasonal- ity e.g. when leaves senescence. We thus hypothesized that the contrast between fruit displays and their background vary throughout the year in a seasonal habitat. If this variation is adaptive, we predicted higher contrasts between fruits and foliage during the fruiting season in a cerrado–savanna vegetation, southeastern Brazil. Based on a six-year data base of fruit ripening and a one-year data set of fruit biomass, we used relectance measurements and contrast analysis to show that fruits with distinct colors difered in the beginning of ripening and the peak of fruit biomass. Black, and particularly red fruits, that have a high contrast against the leaf background, were highly seasonal, peaking in the wet season. Multicolored and yellow fruits were less seasonal, not limited to one season, with a bimodal pattern for yellow ones, represented by two peaks, one in each season. We further supported the hypothesis that seasonal changes in fruit contrasts can be adaptive because fruits contrasted more strongly against their own foliage in the wet season, when most fruits are ripe. Hence, the seasonal variation in fruit colors observed in the cerrado–savanna may be, at least partly, explicable as an adaptation to ensure high conspicuousness to seed dispersers. Traditionally, fruit colors are thought to represent adap- tations to seed dispersers, which are regarded as the main selective pressure on fruit color evolution (Darwin 1862, Kerner 1895, Willson and Whelan 1990). During the 1980s and early 1990s, the inluential ‘syndrome hypothesis’ assumed that seed dispersers have strong preferences for certain colors and that these preferences inluence the evolution of fruit coloration (Janson 1983, Gautier-Hion et al. 1985). Although some studies reported color prefer- ences of fruit consumers (Puckey et al. 1996, Siitari et al. 1999, Whitney 2005), these preferences were often transient (Schmidt and Schaefer 2004), and most studies did not support the syndrome hypothesis owing to inconsistent color choices (Willson et al. 1990, Willson and Comet 1993, Traveset and Willson 1998, Schmidt et al. 2004). Alternatively to color preferences, diurnal seed dispersers might select fruit colors by consistently detecting a color more easily (Kerner 1895). his hypothesis, termed later ‘contrast hypothesis’ (Schmidt et al. 2004) is consistent with the main tenet of signal theory: that detectability is one of the main factors driving the evolution of signals (Schluter and Price 1993). Supporting the contrast hypo- thesis, studies have documented that red and black, the most common colors of bird-dispersed fruits, contrast more strongly against the background than other fruit colors (Lee et al. 1994, Schmidt et al. 2004). Furthermore, the rate at which birds pecked artiicial fruits was a function of the conspicuousness of these fruits in southeastern Brazil (Cazetta et al. 2009). hus, if seed dispersers select fruits mainly on the basis of their detectability, the contrasts between fruit and background (leaves, bark and structures associated with the fruit display) as they are perceived by animals are more important, from an evolutionary perspec- tive, than the color of a fruit per se (Schmidt et al. 2004, Melo et al. 2011). here is also evidence that secondary structures (non-green pedicels, capsules or unripe fruits associated with the ripe fruits, according to Schaefer and Schaefer 2007) increase the contrasts of fruit displays and inluence fruit detection (Willson and Melampy 1983, Burns and Dalen 2002, Schaefer et al. 2007). he contrast between fruit displays and their background are not necessarily constant. In seasonal habitats contrast can change during the year, e.g. when leaves senesce and are shed during the dry season. Burns and Dalen (2002) suggest that temporal changes in the availability of fruits with diferent colors are explicable by changes in fruit conspicuousness against seasonal changes in foliage color- ation. However, Burns and Dalen (2002) characterized fruit Oikos 122: 1335–1342, 2013 doi: 10.1111/j.1600-0706.2013.00328.x © 2013 he Authors. Oikos © 2013 Nordic Society Oikos Subject Editor: Paulo Guimarães Jr. Accepted 21 January 2013