Plant response to the fire regime (1970–2023) in a fynbos World Heritage Site: Ecological indicators for fire management

• Slow-maturing reseeding Proteaceae shrubs are ecological indicators of fynbos fires. • Proteoid recruitment is better after long interval-, warm season-, and larger fires. • During 1970–2023, cool season fires and fire frequency increased in the study area. • Indicator species suggest a minimum tolerable fire return interval of 14 years. • Management should limit cool season fires and fire in young post-fire vegetation. Fire drives ecological processes in most Mediterranean-climate ecosystems of the world, including the fynbos shrublands of the Cape Floristic Region. In these ecosystems, particular plant species may be used as indicators to develop or test thresholds for sound management of fires. We analysed long-term datasets to assess the response of indicator plants (slow-maturing obligatory reseeding Proteaceae shrubs; ‘proteoids’) to the historical fire regime (1970–2023) in the Outeniqua World Heritage Site (OWHS), South Africa. This study constitutes the longest comprehensive analysis of a fire regime in conjunction with plant response in a fynbos protected area to date. This enabled critical assessment of existing rules pertaining to fire management in Mediterranean-climate shrublands, and produced guidelines to inform future management of fire in the southeastern Cape Floristic Region. Assessment of maturation rates and post-fire recruitment of proteoids at OWHS revealed juvenile periods of 5–10 years; superior recruitment success after longer fire return intervals (FRIs), warm-season fires, and larger fires; and long FRIs not generally compromising recruitment. Accordingly, we postulate (i) a minimum tolerable FRI of 14 years, particularly on the drier northern slopes where plants are slower-maturing; (ii) no need for establishing maximum tolerable FRIs; and (iii) a requirement for the predominance of warm-season fires. During the past 54 years, the mean FRI at OWHS was 19–20 years, with 62 % of the area burnt in the warm season (November–April), and only 25 % of the area burnt in fires of natural causes (ignited by lightning or rock falls). However, the mean FRI seemed to have decreased while larger areas have burnt during the cool season in later years. Measured against the responses of indicator species to fires, both these trends constitute reason for concern. Future management interventions should thus be focussed on limiting cool season fires and limiting fires in young post-fire vegetation and in vegetation containing healthy populations of proteoids. Awareness raising among all relevant stakeholders of the impacts of inappropriate fire regimes is furthermore imperative.