Small Island Developing States (SIDS) face multiple developmental challenges including the adverse impacts of climate change. Among these complex challenges is the critical issue of devising strategies and plans to achieve water resource sustainability. The combined effects of hydro-climatic hazards such as droughts, rising sea levels, floods and increasing socio-economic pressures have already begun to adversely impact on SIDS water resources. This review article examined studies on ten English-speaking Caribbean SIDS to explore challenges and opportunities for enhancing water resource sustainability in the Caribbean and to identify emerging research gaps. A desk review and synthesis of existing data and available literature including reports, policy documents, peer-reviewed journal articles and books published over the last ten years were conducted to highlight research gaps in water resource sustainability. The conclusion presents a way forward for SIDS to cope with the consequences of climate change on their vital water resources. The findings from this paper can inform regional polices, strategies and plan and direct research to critical areas where information is needed to support evidenced-based decision making. The review is useful for academics, policymakers and practitioners.

The Republic of Trinidad and Tobago is an archipelagic Small Island Developing State (SIDS), situated on the southern end of the chain of Caribbean islands. Several factors such as climate, topography, and hydrological characteristics increase its susceptibility and vulnerability to flooding which results in adverse socio-economic impacts. Many Caribbean islands, including Trinidad and Tobago lack a flood risk assessment tool which is essential for a proactive mitigation approach to floods, specifically in the Caribbean due to the incommensurate flooding events that occur because of the inherent characteristics of SIDS. This research focuses on the problem of flooding using susceptibility analysis, vulnerability analysis and risk assessment for the island of Trinidad, whilst also presenting a repeatable and appropriate methodology to assess these risks in regions that have similar characteristics to Trinidad. This is especially useful in Caribbean countries because of a lack of internal human capacity to support such efforts. Flood hazard indexes (FHI) and vulnerability indexes (VI) were generated for this study using subjective and objective weighting technique models to identify regions that are affected by flooding. These models were Analytical Hierarchy Process (AHP), Frequency Ratio (FR) and Shannon’s Entropy (SE). Comparative analyses of the three models were conducted to assess the efficacy and accuracy of each to determine which is most suitable. These were used to conduct a risk assessment to identify risks associated with each Regional Corporation of Trinidad. Results indicate that FR is the most accurate weighting technique model to assess flood susceptibility and risk assessment in Trinidad, with an Area Under the Curve (AUC) of 0.76 and 0.64 respectively. This study provides an understanding of the most appropriate weighting techniques that can be used in regions where there are challenges in accessing comprehensive data sets and limitations as it relates to access to advanced technology and technical expertise. The results also provide reasonably accurate outcomes that can assist in identifying priority areas where further quantitative assessments may be required and where mitigation and management efforts should be focused. This is critical for SIDS where vulnerability to flooding is high while access to financial and human resources is limited.

Spatial heterogeneity is a universal phenomenon existing in ecological systems at all scales. In the humid tropics, soil sampling and characterization are complex due to high spatial variability. The inability of researchers to capture this spatial variability inhibits the effective management of resources in terrestrial ecosystems. Hence, simple sampling approaches that are labor and cost-efficient but generate representative samples are required to optimize soil sampling and characterization. In this paper, we used a combination of descriptive statistics and geostatistics to characterize the spatial variability of soil under two land uses. One-way ANOVA and a Student t-test were also used to assess the effectiveness of three distinct sampling designs at capturing the variation in the soil’s physicochemical properties. The wetland soils generally exhibited low spatial variability, whereas the variability of the agroforest ranged from low (coefficient of variance (CV) < 25%) to highly variable (CV > 75%). Results show that in the wetland, CVs of samples obtained using the random sampling design (RSD) and response surface directed design (RSDD) were highly correlated (pearson correlation coefficient (r) = 0.98). Similarly, in the agroforest, the CVs of the stratified random sampling design (SRSD) and RSDD gave almost the same accuracy (r = 0.97). Thus, it is recommended that in this and similar humid-tropical wetland and agroforest ecosystems, the RSD and the RSDD, which require the least samples, can be used, respectively. However, more comprehensive testing is required on a broader range of soils before its more widespread application in other climatic conditions.