Sustainable land use systems planning and management requires a wider understanding of the spatial extent and detailed human-ecosystem interactions astride any landscape. This study assessed the extent of historical, current, and future land use systems in Uganda. The specific objectives were to (i) characterize and assess the extent of historical and current land use systems, and (ii) project future land use systems. The land use systems were defined and classified using spatially explicit land use/cover layers for the years 1990 and 2015, while the future prediction (for the year 2040) was determined using land use systems datasets for both years through a Markov chain model. This study reveals a total of 29 classes of land use systems that can be broadly categorized as follows: three of the land use systems are agricultural, five are under bushland, four under forest, five under grasslands, two under impediments, three under wetlands, five under woodland, one under open water and urban settlement respectively. The highest gains in the land amongst the land use systems were experienced in subsistence agricultural land and grasslands protected, while the highest losses were seen in grasslands unprotected and woodland/forest with low livestock densities. By 2040, subsistence agricultural land is likely to increase by about 1% while tropical high forest with livestock activities is expected to decrease by 0.2%, and woodland/forest unprotected by 0.07%. High demand for agricultural and settlement land are mainly responsible for land use systems patchiness. This study envisages more land degradation and disasters such as landslides, floods, droughts, and so forth to occur in the country, causing more deaths and loss of property, if the rate at which land use systems are expanding is not closely monitored and regulated in the near future.

Long-term trend analysis at local scale for rainfall and temperature is critical for detecting climate change patterns. This study analysed historical (1980–2009), near future (2010–2039), mid- (1940–2069) and end-century (2070–2099) rainfall and temperature over Karamoja sub-region. The Modern Era-Retrospective Analysis for Research and Applications (MERRA) daily climate data provided by the Agricultural Model Inter-comparison and Improvement Project (AgMIP) was used. The AgMIP delta method analysis protocol was used for an ensemble of 20 models under two representative concentration pathways (RCPs 4.5 and 8.5). Historical mean rainfall was 920.1 ± 118.9 mm and minimum, maximum and mean temperature were 16.8 ± 0.5 °C, 30.6 ± 0.4 °C and 32.0 ± 0.7 °C, respectively. Minimum temperature over the historical period significantly rose between 2000 and 2008. Near future rainfall varied by scenario with 1012.9 ± 146.3 mm and 997.5 ± 144.7 mm for RCP4.5 and RCP8.5 respectively; with a sharp rise predicted in 2017. In the mid-century, mean annual rainfall will be 1084.7 ± 137.4 mm and 1205.5 ± 164.9 mm under RCP4.5 and RCP8.5 respectively. The districts of Kaabong and Kotido are projected to experience low rainfall total under RCP4.5 (mid-century) and RCP8.5 (end-century). The minimum temperature is projected to increase by 1.8 °C (RCP4.5) and 2.1 °C (RCP8.5) in mid-century, and by 2.2 °C (RCP4.5) and 4.0 °C (RCP8.5) in end-century.

Uganda continues to experience high-refugee population influx that is beyond the established settlement capacities. Little information is available on the present and future impacts of settling refugees and host-communities on landuse changes and vegetation degradation. This study used Sentinel-2 images (20m) of 2016-2019 and Dinamica Ego-software to determine the current and future spatial areal-extent of changes in 2022. Findings revealed a rarefied increase in areas under subsistence-farming, builtups and refugee-settlements while the losses were seen in savannah-grasslands, wetlands and woodlands. In 2021, the most significant increases in the land will be experienced in builtup-areas, settlements and commercial farming. The most degraded vegetation types were savannah-grassland, woodlands and tree-plantations. These were primarily attributed to unending activities of deforestation, bush-burning, high-refugee population, land-conflicts with host-communities and wetland reclamation. Thus, unless sustainable farming and energy-saving practices are promoted and adopted, the landscape is likely to remain not even with remnants of green-cover.

Despite the achievements reported from using rainwater harvesting systems, the contribution and drawbacks that affect their usage in mountainous landscapes have received little attention. The uptake and usage of domestic rooftop rainwater harvesting systems (RRWHS) in developing countries is on the increase due to increasing water scarcities. We explored the effect of rainfall variability on water supply and the downsides of using the systems by rural households in Uganda. The objectives were to assess the variability of rainfall (1985–2018), categorise RRWHS used, and examine the influence of slope ranges on the placement of systems and also to quantify the harvested and saved rainwater and establish the factors that affected system usage. Rainfall variability was assessed using a Mann–Kendall test, while system contributions and drawbacks were examined using socioeconomic data. A representative of 444 households were selected using a multicluster sampling procedure and interviewed using semistructured questionnaires. Findings revealed that the months of March, April, September, August, and October experienced an upward trend of rainfall with a monthly coefficient of variation between 41 and 126%. With this, households responded by employing fixed (reinforced concrete tanks, corrugated iron tanks, and plastic tanks) and mobile RRWHS (saucepans, metallic drums/plastic drums, jerrycans, and clay pots). At the high altitude, households deployed mostly plastic jerrycans and industrial plastic/metallic drums to harvest and save water. Overall, the mean annual volume of rainwater harvested on the slopes of Mt. Elgon was 163,063 m³/yr, while the potential to save water ranged from 4% to 7% of the annual household water demand. The factors that hindered the deployment of RRWHS to harvest and save water were high operational costs, price fluctuations, unreliable rainfall pattern, inadequate funds, and limited accessibility. The rainfall received if well-harvested and saved can redeem households of water insecurity, though there is an urgent need of subsidies from the government to increase accessibility of the systems.

Agroforestry adoption as a drought adaptation option has an omnibus of opportunities for smallholder farmers in semi-arid regions. This study assessed the severity and frequency of drought and the determinants of agroforestry adoption in Nakasongola District. The episodes were examined using the Standardised Precipitation Index (SPI) set at 3, 6 and 12 months timescales. A cross-sectional survey using semi-structured questionnaires, focus group discussions and key informants were adopted. A total of 200 farmers were randomly selected and studied. The adoption was determined using a binary logistic regression. The SPI results showed that the extreme drought years recorded were 1980, 1984, 1986, 1990, 1995, 1999 and 2000; while the wettest years were 2014, 2012, 2013, 2009 and 2010 as per the 3-time scales. The average return period of severe droughts was 4 years. The levels of agroforestry uptake were higher (85%) between July and June drought period. Agrisilviculture, agrosilvopastoral, silvopastoral and apiculture were the most adopted agroforestry systems by the farmers. The household age, level of education and income were the major significant determinants of agroforestry adoption (p<0.05) in adaptation to drought by the smallholder farmers. The potential benefits of agroforestry adoption included the provision of food, fodder, erosion control and soil fertility enrichment, however, the farmers were mainly constrained by inadequate funds, shortage of tree planting stock, limited extension services and information on agroforestry production. Thus, carrying out massive awareness campaigns on agroforestry practices is more likely to increase the uptake. Key words: Drought, agroforestry, determinants, standardised precipitation index (SPI), adoption, smallholder farmers.