Saltwater intrusion has become a severe issue for the Mekong Delta in Vietnam, especially near the coastline. This issue has led to farmers diversifying from exclusively growing rice to adopting a mixed rice-shrimp system with rice only cultivated in the wet season. However, the nutrient (nitrogen, phosphorus and carbon) cycling and nutrient use efficiency of this system remain poorly understood. To address this knowledge gap, we examined nutrient budgets across 12 farms using integrated rice-shrimp ponds, and in some cases semi-intensive or intensive shrimp grow-out ponds (Penaeus monodon or Penaeus vannamei), over a two-year period (2014-2015). In terms of nutrient budgets, the main nutrient input (92% of the N input, 57% P and 95% C) in the integrated rice-shrimp ponds (IRSPs) came from intake water (excluding C from primary production), while water discharge accounted for the highest output (75% of N output, 41% P, 57% C, excluding C from respiration). The study showed that IRSPs had low dissolved oxygen and high nutrient concentrations which may affect shrimp production. Conversely, salinity levels in the wet season were too high for rice plants thereby affecting rice production. Shrimp survival in the IRSPs was low over the two years (6.3 ± 2.2%), which resulted in the low proportion of nutrients exported from the ponds as harvested shrimp (6% N, 5% P and 10% C). In contrast, the shrimp grow-out ponds (SGOPs), had much higher survival (77.1% for P. vannamei and 59.2% for P. monodon) in three of the six farms where the

The Mekong Delta is the most important rice‐ and shrimp‐producing region for food and economic security in Vietnam. Rice‐shrimp farming is practised where salinity fluctuates substantially between wet and dry seasons. Research points to several potential risk factors for rotational systems, but how these link directly to both rice and shrimp production remains poorly quantified for systems that stock and harvest animals year‐round. We examined water and soil quality of 18 rice‐shrimp‐crab ponds, in which shrimp and crab are grown in both wet and dry seasons, in the Cà Mau Province of Vietnam. Multiple lines of evidence indicated that environmental conditions experienced by both rice and shrimp were suboptimal and contributed to low yields and survival. Year‐round cropping of shrimp and crab was associated with sustained suboptimal salinity, intensified by drought, for the wet‐season cultivation of rice. Although rice seedlings were sown in all 18 ponds, only three had a harvestable crop. Low shrimp production and survival was associated with sustained suboptimal water temperatures (too high), salinity (too high in the dry season and too low in the wet season) and dissolved oxygen concentrations (too low). Food availability and quality may also have affected shrimp production. Improving productivity of rice‐shrimp‐crab ponds in the study region may require (1) separation of rice and shrimp crops and improving efficiency of soil washing practices such that salinity conditions are more suitable for each when grown, and (2) management intervention to increase oxygenation of water, and the availability and quality of food for shrimp.

In recent years, across tropical regions of the world, there has been an expansion of integrated farming systems that combine rice and shrimp production. While these systems were developed as a form of crop-rotation -growing rice in the wet season and shrimp in the dry season -some farmers grow both rice and brackish-water shrimp simultaneously during the wet season. Climatic variability has resulted in considerable crop losses in this system across many regions. Research has yet to identify the complete array of key risk factors, and their potential interactions, for integrated rice-shrimp farming. Consequently, different farming practices and environmental factors that may affect crop production need to be clarified to guide research efforts. We applied a staged, iterative process to develop a probabilistic Bayesian belief network based on expert knowledge that describes the relationships that contribute to the risk of failure of both crops in integrated rice-shrimp farming systems during the wet season. We applied the approach in the Southern Mekong Delta, Vietnam, in the context of a broader research program into the sustainability of the rice-shrimp farming system. The resulting network represents the experts' perceptions of the key risk factors to production and the interactions among them. While both farmers and extension officers contributed to the identification of the processes included in the network, the farmers alone provided estimates of the probability of the relationships among them. The network identified the challenges to minimise the risk of failure for both crops, and the steps farmers can take to mitigate some of them. Overall, farmers perceived they have a better chance to minimise risk of failure for shrimp rather than rice crops, and limited opportunities appear to exist for successful production of both. By engaging the farmers in this process of model development, we were able to identify additional research questions for the broader research team and to identify simple steps the farmers could take to reduce the risk of crop failure.Integrating additional empirical data into this network, as it becomes available, will help Highlights • We converted knowledge from rice-shrimp farmers in the Mekong Delta into a probabilistic model• Key risks to production for both crops revolved around soil and water quality • On average, rice crop failure was perceived to be more likely than for shrimp • Capturing extant agricultural knowledge improves outcomes for research and practice

No abstract

Integrated rice-shrimp ponds (IRSPs) are common in areas of Southeast Asia where saltwater intrudes into rice fields in the dry season, enabling rice production in the wet season, and shrimp farming in the dry season or throughout the year. Previous research has highlighted that IRSPs have periods of low dissolved oxygen concentrations which may have a critical effect on shrimp survival. To understand the causes of low dissolved oxygen, this study examined oxygen fluxes at two IRSPs in Ca Mau Province, Vietnam during a two-year period (two wet seasons and two dry seasons). Sediment oxygen demand (SOD) incubations and whole-system oxygen flux measurements were conducted and compared with a range of water and sediment parameters to explain drivers for low oxygen concentrations. A high percentage of oxygen demand at a whole pond scale was from the sediment; hence SOD drove low oxygen concentrations in the water column. SOD rates were significantly positively correlated with chlorophyll a concentrations in the water column. These findings suggested that algal production in the water column, rather than benthic algal production, or other organic loading, provided an organic carbon source driving SOD. Oxygen demand was much higher than oxygen production within the IRSPs, indicating high bacterial activity and low algal production. This study has shed new light