In South-East Asia, dry-season non-rice crops are becoming more important in tropical lowland rice systems where farmers can access dry-season irrigation. Non-rice crops with higher water productivity than rice, the quotient of yield and water input, provide greater opportunity for improved livelihoods, where dry-season irrigation water supply is scarce, expensive or difficult to access. Paddy soils and paddy fields, however, can limit the water productivity of non-rice crops in rotation with rice in the wet season. In particular, the hardpan formed by puddling during land preparation for rice can seriously limit root-zone depth for the dry-season crop, hence increase drought stress and limit yield.
The broad aim of this thesis was to assess how much current farmer practice of irrigation management involved a mismatch of water supply and plant demand and to assess a number of manageable changes to soil and water management that could improve water productivity and farmer livelihoods. This thesis assesses the ability of three different approaches to improving the water productivity of post-rice dry-season crops in the current lowland rice system, using maize as a test crop. The effect of these techniques on yield, water input and consequent water productivity was assessed in 7 different experiments, in four different locations over three dry seasons, and a greater understanding of the changes in water balance was established. First, the hardpan was disrupted by deep cultivation to try to increase root depth. Second, straw mulch was applied to reduce evaporation and increase yield. Third, water input was reduced in an attempt to better match water supply with demand, either by increasing furrow irrigation interval or using drip irrigation. The water input, yield and water productivity of the observed practice of seeding the maize within small irrigation furrows was also measured. The effect on water input and the yield of either maize or sweet corn was measured with either of these three techniques or a combination of them. In two of these experiments, soil evaporation was measured with micro-lysimeters and transpiration estimated from biomass and vapour pressure deficit, to estimate the effect on water balance components.
This assessment of improved practice was done within the contemporary socio-economic context of Lao PDR. Economic development and better integration with regional and international markets is gradually increasing market opportunities for farm products, but farm labour is becoming more costly and scarce as off-farm and non-farm opportunities increase, irrigation infrastructure and access to mechanization remains limited, and farmers still have very limited access to capital for investment. Hence, the effects of the treatments were assessed not only for their effect on yield and water productivity, but also on labour demand and the magnitude of benefits and costs to the farmer.
The conclusion of this work, was that farmers manage dry-season maize production quite effectively given contemporary constraints in labour, market availability and capital supply. However, there is room for improvement of the timing and method of irrigation, of management of the hardpan in affecting access to soil moisture, and the use of mulch.
Of the two locations tested, hardpan disruption increased yield and water productivity only in the sandy-loam soil at Thassano, Savannakhet province in southern Lao PDR. At Thassano, it increased maize yield and water productivity by 14% in one season with a regular irrigation schedule. However, where the irrigation schedule was less regular in the previous season, it increased the yield with high water input furrow irrigation and with drip irrigation by 24% and 29% respectively, but failed to increase yield with lower water input. Hardpan disruption gave extra yield valued at 3-4 times the extra cost of deep cultivation. At Thassano, there was a negative correlation between hardpan mechanical resistance at seeding and grain yield, so disruption that reduced mechanical resistance a greater amount would likely have increased yield more. Dry-season hardpan disruption did not increase percolation under the following wet-season rice crop in any of the five fields monitored, as the hardpan appeared to reform. Complete hardpan disruption – as represented by mechanical perforation in this thesis – may have increased wet-season percolation, but one or two supplementary irrigation events were usually enough to compensate for the extra percolation, which would likely be reliably available given the system supplies dry season irrigation.
The application of 4 t ha-1 of rice straw mulch did not increase grain yield on a sandy loam at Thassano. At Pakcheng, this mulch increased grain yield in one season by 14%. In the following season, this mulch had no effect on sweet corn with the standard irrigation treatment, but increased fresh ear yield by 42% with low water input before flowering. There was a similar interaction in the same season at nearby Ban Keun Neua, where mulch increased fresh ear yield by 60% with the standard irrigation treatment and 94% with low water input before flowering. Results of two of the experiments show, however, that if straw mulch was applied and water input was reduced before flowering, the gain in water productivity, of 43% and 110%, was much greater than the yield gain. Mulch gave extra benefits 6-16 times greater than the extra cost of applying it, mostly from extra yield and labour savings.
Reducing water input by increasing furrow irrigation interval, to try to better match water supply with demand, did not affect water productivity, as yield declined in proportion to water input. This reduced gross income much more than it reduced water costs with the current charging structure, so it damaged farmers’ livelihoods. Reducing water input with drip irrigation increased water productivity by 52-170%, as yield was similar or higher to standard furrow irrigation. Evaporation from a bare soil was reduced by 50% and percolation was eliminated; hence, water productivity to evapotranspiration of maize in this lowland rice system was estimated at 4.4 kg biomass m-3 and 2.0-2.5 kg grain m-3. Seeding maize in the irrigation furrows, at Muang Kai in Savannakhet, allowed farmers to apply only 140 mm of water to sweet corn and achieve water productivity to water input of 4.4 kg fresh ears m-3, which was comparable to that achieved with drip irrigation elsewhere in the same season.
Hardpan disruption appears a worthwhile technique to improve the livelihoods of farmers growing post-rice dry-season crops on coarser-texture soils. Straw mulch application also seems a worthwhile technique to improve livelihoods, particularly when combined with increased irrigation interval before flowering, although it gave no benefit on a sandy-loam soil typical of southern Lao PDR. In contemporary Lao PDR, however, the value of the treatments from increased yield and reduced labour to weed and irrigate, which allowed more non-farm work, were much greater than the value of saved water with the current water charges.