Convergence among global biogeographical realms in the physiological niche of evergreen and deciduous vegetation

AIM: The seasonal timing and duration of leaf deployment has major implications for earth–atmosphere interactions and biotic processes. While evidence for impacts of global change on the timing of leaf deployment (leaf phenology) is accumulating, knowledge of the mechanisms that determine the duration of leaf deployment (leaf habit) and its geographical distribution remains incomplete. Theoretical models of leaf habit assume that a selective environment leads to convergence to an optimal leaf habit. Here we examine convergence in the physiological niche of different leaf habits. LOCATION: South American, African, Indo‐Pacific and Australian biogeographical realms. METHODS: We delineated the distribution of three phenomes: high leaf area evergreen vegetation, low leaf area evergreen vegetation and deciduous vegetation, using a 1981–2012 time series of normalized difference vegetation index (NDVI) data from the advanced very high resolution radiometer (AVHRR). For each phenome in each of the four biogeographical realms we quantified the physiological niche using a process‐based model of plant growth. The derived physiological niche dimensions of each phenome in each realm were used to project its potential distribution into other, non‐native, realms. Convergence in leaf habit was quantified as the level of consistency among the projections from all realms. RESULTS: The overall agreement between observed and predicted phenomes in the native realm was high (80%), suggesting that the model adequately describes the physiological niches of phenomes. Projections from each phenome into non‐native realms coincided in 89% of the study area, implying widespread convergence of the physiological niches of phenomes. For the remaining 11% of the study area projections of all realms did not coincide, implying that the predicted phenome type is uncertain. MAIN CONCLUSIONS: Modelled physiological niches of the same phenome in different biogeographical realms largely overlap. This implies a dominant role for evolutionary convergence in the biogeography of leaf habit. Overall, our analysis therefore suggests that the impact of global change on the geographical distribution of leaf habit can be predicted by relatively simple physiological models. Potential explanations for non‐convergence on 11% of the study area include historical contingencies and alternative stable states.