Janendra De Costa , Chameesha Madhumali , Nimalka Kankanamge

Climate and its variability strongly influence the structure and dynamics of tropical rainforests, a biome which is critical for regulation of the global climate. We characterized the climate of a series of rainforest plots in Sri Lanka across a wide altitudinal range (117 to 2132 m above sea level) during 1990-2018 and determined its temporal shifts from the climate of 1961-1989. Long-term (1961-2018) climatic data were obtained from WorldClim2 and CRU-TS-4.03 global databases. Soil water deficit was computed, on a monthly (SWD) and cumulative (CSWD) basis, as the difference between rainfall and evapotranspiration using a validated model. During 1990-2018, decreases with altitude were faster in annual mean minimum temperatures (Tmin) than in annual mean maximum temperature (Tmax). The diurnal temperature range (DTR) increased with altitude. Within-year variation patterns of Tmax, Tmin and DTR were different, with peaks in March, April-May, and April respectively. Forests at higher altitudes experienced greater DTRs with greater within-year fluctuation than those at mid- and lower altitudes. Long-term annual rainfall and solar irradiance decreased while SWD and CSWD increased with increasing altitude. All altitudes showed peak SWD and CSWD in February-March. The higher altitudes showed an additional peak in June-July. Inter-annual variability, quantified in terms of the coefficient of variation, was greater for rainfall than for temperature, while CSWD and SWD showed the highest variability. Annual mean Tmax and Tmin increased with time during both periods. Annual total RF decreased with time during 1961-1989, but did show a significant trend during 1990-2018. Consequently, maximum monthly SWD and CSWD decreased from 1961-1989 to 1990-2018. The Dry-Season Index, defined as the annual maximum CSWD, increased during 1961-1989, but decreased during 1990-2018. Altitudinal trends of climatic variables show that the requirement of adaptive mechanisms for climate variability is greatest in montane forests at high altitudes.