Spatial and temporal patterns of mangrove forest canopy gaps at the southern distribution limit

Mangrove forest canopy gaps occur in 35 countries across the global distribution of mangrove forests in some 110 countries. Yet, their spatial and temporal patterns and drivers of their formation and closure remain poorly understood. Here, we investigated whether gaps are distributed randomly, clustered, or dispersed over space and time in two showcase areas in South Africa, uMhlathuze, near Richards Bay, and Beachwood, near Durban, to prime better understanding of the underlying processes. We mapped canopy gaps using free satellite imagery within Google Earth Pro and fixed wing-acquired aerial images and analyzed spatiotemporal patterns using the Ripley’s K function and the Emerging Hot Spot Analysis. Kaplan–Meier analyses estimated the time it takes for gaps to close. Canopy gaps in uMhlathuze occurred spatiotemporally clustered, whereas Beachwood canopy gaps primarily exhibited random spatial patterns. The spatial distribution of canopy gaps was linked to great canopy height at both uMhlathuze and Beachwood, supporting the hypothesis of lightning strikes as causal agents of gap formation. Canopy gaps at uMhlathuze remained open for at least 23 yr, and no gap at Beachwood had closed over the 18-yr study time span, rendering our study region near the southern distribution limit of mangroves a contrast to mangrove forests in many other regions of the world. Without active human intervention, the rising frequency of thunderstorms—and consequently, lightning strikes—is expected to increase canopy gap formation in mangrove forests. This will significantly reduce the ability of mangrove and estuarine systems in South Africa to support climate change mitigation and adaptation, weakening the resilience of coastal socioecological systems.