No two are the same: Assessing variability in broad-leaved savanna tree phenology, with watering, from 2012 to 2014 at Nylsvley, South Africa

Abstract

Ecologists and modellers are yet to fully understand and design accurate models simulating the unpredictable green-up of savanna ecosystems, possibly due to the lack of long-term, phenological monitoring of this biome. In this study, we observed the percentage of leaf age class (new, fully expanded and mature) phenology of two dominant, broad-leaved savanna trees at a weekly scale over the green-up period (August–November) and a monthly scale during the rest of the growing season (December–May) between August 2012 and May 2015 at the Nylsvley Nature Reserve, South Africa. We used an irrigation experiment, commencing in the dry season, to determine whether the simulation of a 20mm rainfall event for several weeks prior to the start of rainfall, could influence the leaf phenology of Burkea africana and Terminalia sericea. Each observed season experienced different environmental conditions from early onset average rainfall (2012), to late-onset high (2013) or low rainfall (2014). Irrigation of T. sericea in 2012 increased green-up rates and in combination with a late season fire during September 2013, resulted in a delayed green-up of the watered trees. Early-greening of B. africana was only observed during the 2014 season. Terminalia sericea showed faster maturation rates than B. africana which indicates the use of a facultative-greening strategy, while B. africana had faster green-up rates after the fire when competition for available nutrients was low. This study has demonstrated that using fine-scale phenological measurements can provide a clearer picture of how trees are greening up in the early stages of the green-up period and showcases the seasonal variability in savanna tree phenology. Investment into long-term phenological monitoring across southern Africa's savannas is needed if we intend to understand cues and predict phenology accurately in the future.