With the increasing world population and accelerated global warming, the main challenge faced today is to develop effective strategies to maintain or increase productivity and prevent environmental deterioration. Rice (Oryza sativa) is the most important agricultural staple for more than half of the world’s population and is critically important for global food security. While rice is essential for ensuring global food security, traditional rice cultivation practised in flooded paddy soils demands unsustainably higher water inputs than other cereal crops, emits a large amount of greenhouse gases, especially CH4, and can accumulate heavy metals such as arsenic (As) in rice grain due to its connection with redox chemistry. In addition, in recent years the renewed interest in application of organic fertilisers, especially as in organically produced rice or amendments e.g. rice straw (produced in huge quantities in intensive cropping systems), can significantly increase CH4 emissions because they add an extra source of carbon substrates for methanogens (Snyder et al., 2009) as well as increase heavy metals in rice grain. Water management practices can alter soil oxygen availability, thereby affecting various processes underlying the production of CH4 and N2O emissions and the uptake of heavy metals such as As in rice grain. The two strategies most often proposed to reduce CH4 emissions are to limit the period of soil submergence (i.e. draining the field) and reduce carbon inputs (through residue management). Alternative irrigation systems such as mid-season drainage and alternate wetting and drying (AWD) irrigation have attracted considerable interest among soil scientists as they can simultaneously save water and reduce CH4 emissions, but their effect on N2O emission and grain yield is less clear. However, irrigation management strategies that reduce yields may simply displace food production and associated GHG emissions to a different region. A deeper understanding of the timing and severity of drainage can shed light on the problem of such a practice. In particular, drainage during the early season (ED) could potentially be effective since the system still contains large amounts of available C from inherent or added soil carbon in the system. Finally, despite promising multiple benefits, the adoption of climate-smart techniques (CSA) by farmers has been observed to be relatively low. Therefore, it is imperative to investigate whether the proposed interventions tested are relevant for farmers and what the determinants are in their choice of adoption of various CSA techniques. Therefore, this thesis was written to understand how alternative water management regimes can be a central force in ensuring multiple agronomic and environmental benefits to help achieve sustainable intensification goals and thus make rice production systems climate-smart. The overall objective of this PhD thesis was to identify and test alternative water-saving management regimes that can reduce GWP from intensified rice farming systems per unit of area as well as per unit of yield, reduce As uptake in grain, and identify the key barriers and determinants of adoption of various available climate-smart technologies for rice ecosystems. The specific research questions of this thesis Summary 150 research were: (i) What impact does the timing and duration of drainage practices (ED, MD, AWD) have on CH4 and N2O emissions? (ii) What is the impact on rice grain yield of the alternative water regime employing ED? (iii) What is the impact of the alternative water regime employing ED on the uptake of important heavy metals (As, Pb, Cd) in rice grain? (iv) How does the water-saving crop establishment with the SWP scheduled irrigation practice affect CH4, N2O and irrigation water use compared with other water-saving alternatives such as AWD? (v) What are the key barriers to the adoption of CSA practices and what are the socio-economic variables that could potentially explain the adoption of CSA techniques in the rice agroecosystem? This thesis contains four research chapters, an introduction (Chapter 1) and a general discussion (Chapter 6). In Chapter 2, a growth chamber experiment was described which tested the effectiveness of seven drainage regimes varying in their timing and duration (combinations of ED and MD) to mitigate CH4 and N2O emissions for 101 days. The results suggested that ED + MD drainage may have the potential to reduce CH4 emissions and yield-scaled GWP by 85-90 % compared with CF, and by 75-77 % compared with MD only. A combination of (short or long) ED drainage and one MD drainage episode was found to be the most effective in mitigating CH4 emissions without negatively affecting yield. In particular, compared with CF, the long early-season drainage treatments LE+SM and LE+LM significantly decreased yield-scaled GWP by 85 % and 87 % respectively. Overall N2O emissions were small and not significantly affected by ED. It is concluded that ED+MD drainage could be an effective low-tech option for small-scale farmers to reduce GHG emissions and save water while maintaining yield. Chapter 3 described a field experiment that was set up at the Experimental Station of the International Rice Research Institute (IRRI) in Los Baños in the Philippines to answer Research Questions 1, 2 and 3. This tested the effect of early AWD (e-AWD) vs. continuous flooding (CF) water management practices on grain yields, GHG emissions and grain arsenic levels in a split-plot field experiment with organic fertilisers. The treatments included i) farmyard manure, ii) compost and iii) biogas digestate, alone or in combination with mineral fertiliser. The e-AWD water regime demonstrated no difference in yield for the organic treatments. Yields significantly increased by 5-16 % in the combination treatments. The e-AWD water regime reduced seasonal CH4 emissions by 71-85 % for organic treatments and by 51-76 % for combination treatments; this was linked to a reduction in dissolved organic carbon (DOC) of 15-47 %, thereby reducing methanogenesis. N2O emissions increased by 23-305 %, but accounted for <20 % of the global warming potential (GWP). Area and yield-scaled GWP were reduced by 67-83 %. The e-AWD altered soil redox potentials, which resulted in a reduction in grain arsenic and lead concentrations of up to 66 and 73% respectively. Application of organic fertiliser and a mild drainage regime such as e-AWD also facilitated Summary 151 the reduction of grain cadmium levels by up to 33 % in organic systems. Structural equation modelling revealed that DOC, redox, ammonium and root biomass were the key traits regulating emissions and maintaining yield. These results suggested that in flooded rice systems receiving organic amendments or organic management, the e-AWD water regime can achieve multiple environmental and food safety objectives without compromising yield. Chapter 4 described the second field experiment, which was conducted during the dry season at the International Rice Research Institute (IRRI) in Los Baños, Laguna in the Philippines. The aim of this study was to evaluate the impact of water-saving irrigation (alternate wetting and drying (AWD) vs. soil water potential (SWP)), contrasting land establishment (puddling vs. reduced tillage) and fertiliser application methods (broadcast vs. liquid fertilisation) on water-use efficiency, GHG emissions and rice yield. The experiment was laid out in a randomised complete block design with eight treatments (all combinations of the three factors) and four replicates. AWD combined with broadcasting fertilisation was superior to SWP in terms of maintaining yield. However, seasonal nitrous oxide (N2O) emissions were significantly reduced by 64 % and 66 % in the Broadcast-SWP and Liquid fertiliser-SWP treatment, respectively compared with corresponding treatments in AWD. The SWP also significantly reduced seasonal methane (CH4) emissions by 34 and 30 % in the broadcast and liquid fertilisation treatments, respectively. Area-scaled GWPs were reduced by 48 % and 54 % in the Broadcast-SWP and Liquid fertiliser-SWP treatments respectively compared with the corresponding treatments in AWD. Compared with AWD, the broadcast and liquid fertilisation in SWP irrigation treatments reduced yield-scaled GWPs by 46 % and 37 % respectively. In terms of suitability and based on yield-scaled GWPs, the treatments can be ordered as follows: Broadcast-SWP < Broadcast-AWD = Liquid fertiliser-SWP < Liquid fertiliser-AWD. Growing-season water use was 15 % lower in the SWP treatments compared with the water-saving AWD. Reduced tillage reduced additional water use during land preparation. The conclusions of this study were that improved water management and timely coordination of N fertiliser application with crop demand can reduce water use, N loss via N2O emissions, and CH4 emissions. Chapter 5 described a farm-household survey in Philippines conducted in collaboration with the International Rice Research Institute (IRRI) to assess how the adoption of climate-smart technologies is dependent on the local socio-economic and biophysical environments, food security and gender. A total of 250 households from five municipalities in the province of Bulacan in Central Luzon were surveyed. The results showed that households in which the household head is young and educated are more likely to adopt climate mitigation techniques. A unit increase in the number of years of schooling would result in a 1.5 % increase in the probability of adoption of mitigation and a 1% increase in the probability of adoption of Summary 152 adaptation techniques, as well as a 0.6 % increase in the probability of adoption of good practices. Male-headed households were more likely to adopt CSA technologies than female-headed households, which is linked to the women’s low resource endowments as well as their more limited contact with extension technicians and research institutes promoting such technologies. Access to extension services increased the probability of adoption of climate mitigation, adaptation and good practices by 19, 17 and 15 % respectively. Adoption of climate mitigation techniques and good practices were most significantly impacted by the connection with research institutes, and resulted in a 21 and 15 % increase in the adoption respectively. Similarly, farmer-to-farmer extension, which acts as social capital, increased the probability of adoption of climate mitigation, adaptations and good practices by 16, 9 and 12 % respectively. Households that are food secure were 11 and 21 % more likely to adopt the good practice and climate change mitigation practices respectively than those that were food insecure. Again, increasing women’s empowerment increased the likelihood of the adoption of climate adaptation practice by 23 %. It was concluded that in order to increase adoption of CSA technologies, government investment in sustainable irrigation and the provision of funds and assistance for support programmes are required, as well as the targeting of young farmers and female farmers in terms of improving their access to various information channels. The general discussion (Chapter 6) addressed the main findings of the research and their implications in a broader context. The thesis demonstrates that early-season drainage employed in combination with midseason drainage or AWD has great potential to reduce the total GHG budget from rice paddy systems without affecting yield. Moreover, the e-AWD water regime was also found to be a very effective strategy to reduce water use and uptake of grain heavy metal such as arsenic (by up to 66 %), lead (by up to 73 %) and cadmium (33 %) in organically fertilised rice. Irrigation scheduling informed by soil water potential (SWP) was found to be a more effective water-saving and GHG mitigation strategy than the AWD practice with the equivalent rice yield, but was limited by the requirement for initial investment and by being knowledge intensive. Non-puddling reduced tillage land preparation instead of puddling can reduce additional water use, labour and energy. Water scarcity, lack of capital and shortage of good quality seeds were found to be the most limiting factors for rice-growing farmers in the Philippines. The gender, age and education of farmers were found to have a significant effect on the adoption of different categories of CSA technologies. Contact with research institutes was found to have a significant impact on the adoption of CSA technologies. To increase adoption, young farmers and female farmers need to be specifically targeted in terms of government support, e.g. with incentives and better access to various information channels. Finally, the perspectives and future research directions section offers inspiration and guidance for further research to future science enthusiasts. The outcomes presented in this thesis demonstrate that e-AWD is a Summary 153 promising management practice that offers multiple environmental benefits. However, due to limitations of time and resources, long-term field trials could not be established. These should be carried out to realise the benefits of alternative water regimes in different regions and identify soil properties that can affect e-AWD performance, highlighting the need for regional research to fine-tune recommendations for farmers.