Abstract

AbstractAimFlower traits are critical to the mutualistic networks that underpin ecosystem function. However, the broad‐scale drivers of flower traits are unclear. Using the megadiverse islands of Malesia, we test three hypotheses: (1) Flower size, openness and colourfulness (i.e. the trend from white to colourful flowers) will decrease with increasing temperature, but not with increasing precipitation seasonality. (2) Flower size, openness and colourfulness will increase with island area, and decrease with isolation. (3) Models parameterised by climate, rather than non‐climatic island characteristics will have the greatest predictive capacity as climate affects both pollinator communities and energetic/physiological constraints upon flower traits.LocationFourteen Malesian islands.TaxonAngiosperms.MethodsWe applied Bayesian phylogenetic models of occurrence for >2500 species in 400 m elevation bands, across islands, as a function of flower trait interactions with either climatic variables or islands characteristics. We compare predictive capacity of models based on climate versus island characteristics using Leave One Out‐Cross Validation.ResultsAll six flower traits examined varied with at least temperature, precipitation seasonality, island area and isolation. Flowers are larger in cold environments. The lowland tropics are mostly inhabited by white flowered species. At cold high elevations, red and pink flowered species are more frequent whereas green and purple flowered species increase in the drought‐prone seasonally dry tropics. Flower openness declined with precipitation seasonality and increased with isolation. Models parameterised by climate performed best for perianth length, red and white flowers, whereas pink flowers and flower openness were better predicted using island characteristics.Main ConclusionsThere are relationships between flower traits and climatic gradients within Malesia. However, island characteristics have also left a legacy upon current flower trait distributions. These island biogeographical impacts are likely key for the eco‐evolutionary drivers of flower traits.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call