Eva Marinela Kasinda

Fire is a key ecological process shaping savanna vegetation, yet its effects on tree survival may vary with fire season, tree size, and species identity. This study evaluated the effects of seasonal fire on tree survival, biodiversity, and species composition at the Makambu Fire Experiment in the Kavango West Region of Namibia. Tree survival was assessed across four treatments: Control (unburned), Early Dry Season (EDS), Late Dry Season (LDS), and Late Wet Season (LWS), using a randomized complete block design comprising 20 experimental plots. A total of 1,197 trees representing nine woody species were assessed, including 1,054 live and 143 dead trees. A generalized linear mixed-effects model evaluated the effects of fire treatment, tree height, DBH, and species identity on survival, with plot included as a random intercept. Biodiversity metrics were compared among treatments using Kruskal–Wallis tests, while species composition was assessed using Bray–Curtis dissimilarities, non-metric multidimensional scaling (NMDS), and PERMANOVA. Survival was 88.9% in burned plots and 86.2% in unburned plots. Tree height was positively associated with survival, whereas DBH showed a negative conditional association with survival after accounting for other predictors. Species identity also significantly influenced survival, while fire treatment did not. Species richness, Shannon diversity, Simpson diversity, evenness, and dominance did not differ significantly among fire treatments. Similarly, species composition did not differ significantly among treatments. Overall, tree survival was more strongly associated with tree structural characteristics and species identity than with fire season, while seasonal burning did not produce detectable differences in community-level biodiversity or tree species composition under the conditions of this experiment. These findings suggest considerable resilience of the woody-tree community to the seasonal fire regimes examined, although continued long-term monitoring is needed to determine whether repeated fire effects become evident through changes in tree structure, recruitment, and regeneration.

Fire is a key ecological driver in savanna systems, influencing soil biogeochemical processes by releasing nutrients but also causing potential losses through volatilization and leaching. In this study, we investigated how fire seasonality and soil depth influence both Mehlich-3 extractable nutrients and total elemental concentrations. This study was conducted in a long-term fire experiment consisting of 20 plots subjected to seasonal burning treatments (dry season, late dry season, late wet season, and an unburned control). Soil samples were collected from two different depth horizons (0–15 cm and 15–30 cm) and analyzed for nutrient concentrations. We used two-way PERMANOVA, Scheirer–Ray–Hare tests, and non-metric multidimensional scaling (NMDS) to assess treatment effects and interactions. Results showed that fire seasonality did not significantly affect Mehlich-3 extractable nutrients; in contrast, soil depth strongly structured nutrient profiles, with higher concentrations in the upper soil layer, while total elemental concentrations were consistently enriched in the deeper soil horizon. Fire seasonality influenced total elemental nutrient pools, with late dry season burns enhancing potassium and other cations, likely due to combustion of accumulated dry biomass and subsequent ash deposition. Depth effects were particularly pronounced for TN and OC, which were elevated in deeper layers due to the presence of mineral-associated organic matter OM. Our findings highlight that vertical stratification is the dominant control of nutrient distribution, while fire seasonality modulates specific nutrient dynamics. These results underscore the need to incorporate soil profile depth and fire timing when evaluating post-fire soil fertility and ecosystem resilience in savanna landscapes.