Effects of Spatio-Temporal and Tidal Height Variability on Soil CO2 and CH4 Fluxes in Mangroves of the Lamu Archipelago, Kenya

This study quantified the influence of temperature, atmospheric relative humidity, and seasonal and tidal variability on soil CO2 and CH4 fluxes in mangrove ecosystems on the Southern Swamp and the Pate Islands of the Lamu Archipelago, Kenya. The field measurements were conducted during wet and dry seasons and at varying tidal heights for 2 consecutive years (2023 and 2024). Because measurements were repeatedly collected from permanent plots, mixed-effects models accounting for the repeated sampling structure were used as the primary inferential analyses, while Spearman’s rank correlations and Mann–Whitney U were used to characterize unadjusted bivariate associations and group differences. In the unadjusted analysis, CO2 fluxes were significantly higher during the dry season than during the wet season (median 3.72 versus 1.91 µmol CO2 m−2 s−1; p = 0.020). However, season was not a significant independent predictor of CO2 flux in the adjusted mixed-effects model. Similarly, CH4 flux showed a significant positive bivariate association with temperature and a significant negative association with tidal height. In the adjusted mixed-effects model, temperature remained significantly associated with CH4 flux (β = 2.961, p = 0.045, 95% CI: 0.410–6.064), whereas tidal height was no longer significant. Atmospheric relative humidity did not show a significant independent effect on either gas, and mangrove block showed no significant independent main effect. Overall, the findings indicate that temperature was the principal measured environmental predictor independently associated with CH4 flux, whereas the observed dry-season increase in CO2 flux and negative tidal-height association with CH4 represented unadjusted patterns that were not retained as significant independent main effects after accounting for covariates and repeated measurements. These results underscore the importance of distinguishing exploratory bivariate patterns from adjusted effects when assessing environmental controls on greenhouse gas fluxes in mangrove ecosystems.