Fruiting season of many Sri Lankan tropical montane species is not synchronised and may not occur when conditions are favourable for seedling establishment. We hypothesised that species with different fruiting seasons have different seed dormancy mechanisms to synchronise timing of germination with a favourable season for establishment. Using six species with different fruiting seasons, we tested this hypothesis. Germination and imbibition of intact and manually scarified seeds were studied. Effect of GA on germination was examined. Embryo length:seed length (E:S) ratio of freshly matured seeds and of those with a split seed coat was determined. Time taken for radicle and plumule emergence and morphological changes of the embryos were recorded. The radicle emerged from Ardisia missionis, Bheza nitidissima and Gaetnera walkeri seeds within 30 days, whereas it took >30 days in other species. Embryos grew in seeds of B. nitidissima and G. walkeri prior to radicle emergence but not in Microtropis wallichiana, Nothapodytes nimmoniana and Symplocos cochinchinensis. A considerable delay was observed between radicle and plumule emergence in all six species. Warm stratification and/or GA promoted germination of all species. All the tested species have epicotyl dormancy. Seeds of B. nitidissima and G. walkeri have non-deep simple morphophysiological epicotyl dormancy, and the other four species have non-deep physiological epicotyl dormancy. Differences in radicle and epicotyl dormancy promote synchronisation of germination to a favourable time for seedling development. Therefore, information on dormancy-breaking and germination requirements of both radicle and epicotyl are needed to determine the kind of dormancy of a particular species.

River regulation as a result of construction of dams caused global-scale ecological changes to riverine ecosystems. Studies have shown that vegetation alterations may take place due to changes in the water flow, flooding frequencies, and fluctuating water table associated due to dam construction. The Accelerated Mahaweli Project (AMP) which started in 1970's involved construction of five major dams across the Mahaweli River. However, no study has been undertaken so far to assess the dam-induced impacts on downstream riverine forests. The present study was aimed to investigate any dam-induced impacts on the riverine vegetation in the immediate downstream areas of the AMP. In order to achieve this objective, a stretch of Mahaweli River with a regulated water flow (in Minipe area) and a tributary of Mahaweli river with an unregulated water flow Badulu Oya was selected for the study. A complete vegetation sampling was carried out using eight 10 m x 30 m belt transects laid perpendicular to the river flow. Each belt transect was divided into three 10m x 10m zones (hereafter named as lower, mid and upper slopes). The vegetation was quantified under three categories canopy vegetation, understory and ground vegetation. Paired t-test (95% CI) in Minitab 16.0 revealed that the canopy vegetation abundance in the lower slope is significantly higher (p ≤ 0.039) at Badulu Oya than that of Minipe. However, no significant difference was observed in other two zones in between the two rivers. The lower slope showed similar trend in terms of canopy vegetation diversity though it was not significant. Terminalia arjuna and Ficus racemosa dominated the canopy vegetation in Badulu Oya and Minipe, respectively. Both abundance and diversity of the understory vegetation was significantly higher at Badulu Oya than that of Minipe and this effect is more prominent in the lower-and mid-slopes than in the upper slope. Dimorphocalyx glabellus dominated the understory vegetation at Badulu Oya while at Minipe it was Cipadessa baccifera. Riverine vegetation in both rivers was dominated by native species, however no significant difference observed in the abundance of invasive plants. Interestingly, at Badulu Oya the abundance of the riverine vegetation decreased towards the upper slope but at Minipe no such trend was observed. The preliminary findings of the research indicate that the regulation of rivers through damming can have long-term impacts on the riverine vegetation by changing its composition and diversity. However, further studies will authenticate this inference.

Abstract:Since the initiation of large-scale development in late 1970s, the Mahaweli River basin in Sri Lanka has experienced significant changes. However, no comprehensive study has been undertaken so far to evaluate the impacts of river regulation on associated ecosystems including floodplains in the downstream. The present study was aimed at identifying the impacts due to both river regulation and other anthropogenic activities on inland floodplain habitats (locally known as villus) located along the final stretch of the River Mahaweli before reaching the Indian Ocean. Four villus, Handapana (HAN), Bendiya (BEN), Karapola (KAR) and Gengala (GEN), were selected for the study. HAN and BEN can be considered as highly influenced (HI) by river regulation while KAR and GEN as less influenced (LI) due to their respective locations. Due to the absence of pre-regulation vegetation data, HI villus were compared with LI villus in order to explore any potential impacts of river regulation. Vegetation was enumerated using belt transect method. To find out other on-going anthropogenic impacts on these villu ecosystems, a survey was conducted using 100 individuals living in two villages located nearby. The results revealed some significant modification in the composition and the diversity of the vegetation, most possibly due to river regulation and other on-going anthropogenic activities. However, the most notable changes were recorded in the herbaceous layer. Some native aquatic herbaceous species have been completely absent over the period of two decades since the developmental activities begun, while some exotic invasive aquatic species (Eichhornia crassipes) dominated the herbaceous layer in HI villus threatening the survival of the remaining native species. Density and richness of lianas too diminished significantly in HI villus perhaps due to changes of micro-habitat conditions as a result of river regulation and also due to over-harvesting for commercial purposes. The results suggest that these ecosystems have been altered over the years due to culmination of factors including altered flow regimes following river regulation and some on-going human influences. The present study highlights the importance of regulating such human influences on villus including fishing and extracting cane and reed in order to protect these vulnerable ecosystems for future generations. The potential of these ecosystems to develop ecotourism has also been emphasized.