An adaptive cycle hypothesis of semi-arid floodplain vegetation productivity in dry and wet resource states

Abstract

Spatial and temporal variability in flooding plays a significant role in the productivity of semi-arid floodplain ecosystems. Floodplains may be perceived as boom-bust systems, but this model does not account for transitions that may occur between wet and dry floodplain states. This study used the concept of adaptive cycles to examine how floodplain vegetation productivity changes in response to wetting and drying. Floodplain vegetation productivity was tracked through a wet and dry state using the normalized difference vegetation index (NDVI). Floodplain inundation revealed complex vegetation productivity responses to resource availability. There was low NDVI in the dry phase, whereas vegetation vigour increased and decreased through the wetting, wet and drying phases. There was a marked difference in NDVI class area, number of transitions, direction of transitions, probability of transitions and NDVI class diversity between the dry phase and the combined wetting, wet and drying phases of floodplain inundation. The distribution of transition probabilities was platykurtic in the dry phase and bimodal during the wetting, wet and drying phases. Overall, anti-clockwise hysteresis was the dominant direction of hysteresis. All vegetation productivity measures demonstrated a switch in direction during the wet phase. The hysteresis observed in this study indicates the cyclic nature of vegetation response to floodplain inundation through dry, wetting, wet and drying phases. We propose that vegetation productivity response follows an adaptive cycle and that this is an appropriate model for understanding the complexity of semi-arid floodplain vegetation response to wetting and drying.