Vegetative and reproductive phenology in a tropical grassland-savanna-forest gradient

Question Climate, particularly precipitation, is the primary factor driving tropical plant phenology because of its relevant role in regulating water availability. However, differences in soil physicochemical properties can also drive resource availability, potentially affecting plant phenology, especially under similar rainfall regime environments. We determined whether seasonality and synchrony of vegetative and reproductive phenophases differ along a grassland-savanna-forest gradient under the same climate regime in a Neotropical seasonal ecosystem. We hypothesized that species growing on shallow, nutrient-impoverished soils have a higher capacity to respond to precipitation pulses than rich soils. Location Serra do Cipó, southeastern Brazil Methods We quantified soil physicochemical properties associated with fertility and water retention of the grassland (campo rupestre), savanna (cerrado) and forest, and determined phenological strategies of 70 species across the three vegetations types by monitoring vegetative and reproductive phenophases over one year. For each phenophase, in each vegetation type, we evaluated seasonality and synchrony of phenological patterns. We quantified the phylogenetic signal for phenophases to disentangle the relative roles of historical versus ecological drivers of plant phenology. Results Soils from campo rupestre were more nutrient- and water-limited than those of cerrado and forest. Cerrado and forest communities had similar phenological strategies, whereas species from campo rupestre showed strategies that maximize resource acquisition and conservation. In the cerrado and forest, leafing and flowering patterns were seasonal, with leaf flushing and flowering peak at the onset of the rainy season and leaf senescence in the dry season. Conversely, species from campo rupestre showed continuous leafing and flowering patterns, while fruiting was seasonal. Phenophases did not show phylogenetic signal. Conclusions We demonstrated changes in vegetative and reproductive phenology that are likely associated with variation in soil physicochemical properties. We argue that more severe edaphic filters may have shaped different phenological patterns in campo rupestre.