Legumes and reduced tillage for rice and maize based cropping in the Democratic Peoples Republic of Korea (DPRK)

Agreement between the NSW Agriculture (Commissioned Organisation) and ACIAR for this project was established on the 13 March 2002. Two appropriate locations for experimental work in the DPRK were agreed to with DPRK authorities, Hyongsan Cooperative Farm and Unsong Cooperative Farm, and at each location a rice-legume and a maize-legume rotation experiment have been established, in April 2002. These, and sites for soil conservation studies were inspected by two Australian scientists during a visit in May 2002. The trials had been set out largely in accordance with previously forwarded design documents, and essential inputs of superphosphate for the trials were received into DPRK from China. The rotations commenced with the summer cereal phase, which was harvested in September 2002. Two Australian scientists visited the experimental sites in September 2002. Rice crops were showing visual responses to N fertiliser treatments, indicating soil N deficiency and portending cereal responses to legumes. Maize crops were not showing symptoms of N deficiency (nor in an earlier visit by Dr Tony Fischer in July); however we subsequently learnt that a recommended application of potash had not taken place, which may have obviated response to N fertiliser. The scientists noted the failure in the maize rotations of the summer intercrop legume treatment based on clover, due to excessive shading by maize. Alternatives are being sought, and relevant to this, the successful re-establishment of self-sown vetch under maize in a sub-trial was encouraging. The winter legume phase involving vetch commenced in September 2002; the scientists were able to witness sowing of the earliest of these treatments: the broadcasting of vetch into standing rice. In the case of maize, a sub-trial in which vetch was established successfully from broadcasting in August, prior to harvesting maize, was encouraging.

At each farm a soil conservation / tillage study has been established in upland maize country. These also commenced with the sowing of maize in June 2002. Seedling establishment was improved 25% in the direct-drilled treatment compared with the traditional cultivation treatment.

Capacity-building material inputs, including field equipment and consumables such as seed, fertiliser, sampling bags; laboratory consumables and hardware supplies; office supplies and computers for data storage and analysis were sent to the DPRK mostly by sea freight but also by air for urgent needs. These were received in November 2002. Sowing and mulching equipment were designed and manufactured at the Agricultural Engineering Dept. of China Agricultural University and along with bicycles for short distance transport to trials, were received in the DPRK in December 2002. Promised financial assistance was successfully forwarded to the DPRK project coordinator.

Capacity building in skills has involved the successful practical study tour by two DPRK scientists to NSW Agriculture (Wagga Wagga and Gosford), Charles Sturt University, and Rutherglen Research Institute in October 2002. These scientists successfully completed tasks in soil microbiology and chemistry mostly related to legume production and nitrogen cycling in agricultural systems. The scientists also retrieved much scientific literature.

The Australian field trial was established as per plan but, subsequently, had to be abandoned because of the severe drought in 2002. It will recommence in 2003. Financial 'savings' from not executing this trial in 2002 were re-directed to purchasing supplies promised for the DPRK in 2003 and 2004. This also avoided the need for further expenditure on container freightage.

With the exception of the unavoidable re-siting of two experiments in the DPRK and postponement of the Australian trial, the project has progressed according to plan. However, we have yet to receive soil and production data from the DPRK trials. We are presently taking steps to determine the cause of the delay.

Field-based experimental studies in the Democratic Peoples Republic of Korea (DPRK), aimed at increasing rice and maize yields and reducing soil erosion, are continuing following their establishment in 2002. DPRK authorities are cooperating with commitment to the conduct of the studies and the release of experimental data. The export/sharing of project data is a breakthrough for work of this type in DPRK, exemplifying the trust that has developed between the scientific institutions and scientists in DPRK and Australia.

The introduction of hairy vetch into maize and rice cropping systems, as an autumn-spring legume phase to increase soil fertility and organic matter has been well received. The first results on the productivity of hairy vetch are indicating a potential for up to 24 tonnes of fresh weight production, which is feasibly an input of 100 kg of nitrogen to the system. However, water-logging of the vetch, caused by pondage of snow melt, has produced large variability in hairy vetch productivity and solutions are being sought to address this problem. The potential yields of hairy vetch sown pre-winter substantially exceeded those of common vetch (up to 10 tonne of fresh weight) established in early spring, so hairy vetch is likely to have a greater impact than common vetch in restoring soil fertility and organic matter. The maize and rice production on the experimental sites, at both of the cooperative farms involved in the study, were responsive to application of urea fertiliser. Grain yield increases of up to 1.4 t/ha in the rice production experiments and up to 2.8 t/ha in the maize production experiments, in response to urea application and with adequate plant available phosphorus from superphosphate application, indicated the inherent nitrogen infertility in the soils. Moreover, a preliminary analysis of the soil analytical results has revealed that responses to urea may be limited by an insufficiency of potassium. On the basis of the first results for the cereal crop responses to vetch, the legume increased grain yield production: by 0.6 to 0.8 t/ha in the rice system, and 0.4 to 0.8 t/ha in the maize system (or about 90% of the response to 136 kg of urea per hectare). Therefore, hairy vetch is likely to play a key role in reducing the agricultural need for urea fertiliser, which is largely acquired as aid, and ultimately in helping to restore rice and maize yields to levels that will make the DPRK self sufficient for staple grain production.

An associated study is also evaluating the potential for alternative legumes to vetch, to cater for the range of climates, soil types and agricultural niches in DPRK. In the first of these studies, a trial was conducted in 2002/03 to evaluate437 genotypes of legume in terms of their ability to survive the winter survival and produce biomass in spring. This experiment highlighted the superior performance of 25 genotypes, mostly from the Medicago genus.

The soil conservation study is also well under-way. This study is vital to the eventual stability of the maize cropping system. Early work is evaluating the potential for growing maize as a direct-drilled crop into stubble comprising vetch and maize residues. The first of the results on the effects of conservation soil treatments on run-off and sediment loss have been released. They are particularly encouraging, with losses of eroded soil during high intensity rainfall reduced by 60-75% as a result of growing maize on beds and retaining maize stubble compared with maize grown in the traditional manner using hilled rows and nil stubble retention.

Capacity building is continuing. A wide range of laboratory and field consumables and equipment continue to be supplied and these items are helping to maintain the field trials and to collect the data essential to understanding the performance of the crop and soil treatments. In particular, the DPRK project scientists have now been supplied with computers and software necessary to storing and manipulating experimental data. During 2003, capacity building was provided through a visit to DPRK by Professor Ted Wolfe (agronomy, data handling) and Dr Phillip Eberbach (tillage, soil measurements, run-off) from Charles Sturt University. Unfortunately the outbreak of SARS virus limited visitations by other project scientists in 2003.

The Australian field trial was re-initiated in 2003 following an interruption due to drought in 2002. Productivity and development data on a mixture of aerial seeding, erect clovers, and associated soil nitrogen, moisture and acidity data, have been collected to provide a greater understanding of the role of an annual forage conservation legume crop in stimulating wheat production and sustaining soil fertility.

In April 2004 the mid-term review of the DPRK component of the project was carried out in Pyongyang, DPRK. This review workshopped results and progress over 2002 and 2003. Comparative yields of rice and maize between these years showed that one of the subduing influences on grain yields is rain and cloudy weather during July-August, which is capable of reducing yields (2003) by 30-40%. This analysis justified the need for several seasons of experiments in order to correctly determine the impact of experimental treatments.

The complete set of data available to the review confirmed that a winter vetch green manure improved 2003 rice yields by 20%, or the equivalent of an addition of 60 kgN/ha as urea. Maize yield response to winter vetch, 11%, was less than in the rice system however, the 2003 response was adversely influenced by removal of vetch by grazing animals and by historically high N fertility at one site.

The rice and maize rotation and upland soil conservation field experiments have been sustained in 2004. Whilst we are yet to receive comprehensive data for 2004, we visited DPRK again in October and obtained interim results. In the rice rotation winter vetch yield was again highly variable(4 to 36 t fresh weight/ha), resulting from water logging in low lying rice bays (similar to 2003). Also similar to 2003, yield response of rice to vetch was of the order of 15-20%, again equivalent to at least 60 kgN/ha as applied urea. During that visit we leant also that too much N, as with treatments involving vetch combined with urea (60 kgN/ha) or a high rate of urea (100 kgN/ha), caused bacterial wilt in short-season rice and reduced yields. Therefore, greater understanding will need to be gained of the N cycled from vetch, fertiliser and soil sources, interactive with rice variety, before producing an optimal management package for legume application to rice production.

In the upland maize system yields of spring-sown winter vetch in 2004 were 14 to 21 t fresh weight / ha; vetch sown between maize rows yielded a further 3 to 5 t fresh weight/ha. At one site (data not yet available at the second site), maize was said to have increased its yield by 1.9 t/ha (50%) in response to vetch; equivalent to a yield with 60 kgN/ha as urea.

At the mid-term review meeting, 2003 data on sediment loss in the soil conservation experiments was presented. These data showed that culturing maize on flat beds together with retaining surface stubbles of the winter vetch and about 2 t/ha of maize stubble, reduced sediment loss to 26 to 37% of that which occurred with the conventional maize culture system; ie., ridges with no stubble. This reduction was directly associated with lower run-off of water during intense summer storms. Run-off and sediment data for 2004 had not been compiled at the time of our visit in October 2004. However, the 2004 maize yield data in the tillage experiments showed that yield was similar for maize cultured on either ridges or beds, so that the soil conserving system would not result in a penalty on yield.

At the time of our October visit, the 2004/05 winter vetch had been established and was emerging. It is intended to establish cereal crops in 2005 so that the project will have completed 3 legume-cereal cycles.

Capacity building in 2004 was primarily our delivering an experimental design and statistical analysis course during the October visit. This was received enthusiastically.

In the Australian study, the legume, soil and climatic data collected in 2003 were subjected to the WNMM model (cropping systems model: University of Melbourne). The simulations of legume biomass, soil nitrate and water, were sufficiently encouraging to continue with the application and refinement of this model. In 2004, wheat was established on the legume treatments imposed in 2003. Sufficient crop, soil and water data have been collected from the 2004 wheat season to test the CERES or SUCROS models for reliability in predicting wheat performance following annual legume crops.

Year 2005 saw the final harvests of rice and maize in cropping systems experiments in DPRK. The impact of sequential seasonal cropping of these cereals with a green manure legume (vetch) was assessed over 3 years. In 2005, the green manure legume, established in autumn (September) 2004, produced poorly as a result of severe cold spell in March 2005. This poor growth probably reduced the yield responses of the rice and maize to the legume, which at the 2 Farms supporting the trials were 0.25 & 0.84 t/ha for rice, and 1.67 & 1.08 t/ha for maize, relative to yields from the traditional low input rice or maize farming practice. Nevertheless, rice yields in the legume-rice rotation were 90-97% of the yield of rice fertilised with 60 kg N/ha as urea; and maize yields in the legume-maize rotation reached 80-114% of maize fertilised with the same rate of urea. Hence, the regular inclusion of an autumn-spring legume phase in rice or maize cropping systems is helping to sustain higher yields of these cereals. Widely adopted, the legume-based rice and maize production systems would help to overcome the usual annual deficit in grain products that characterises food insecurity in the DPRK. As well, green manure legumes would help to reduce the dependence of cereal production on urea, thereby reducing dependence on fertiliser aid.

Hairy vetch (Vicia villosa) is tolerant of the low winter temperatures in DPRK but it is vulnerable to severely cold winters or cold snaps, as occurred in 2004/05. Legume selection studies carried out during this severe winter identified lines of three species of Medicago (M. sativa, M. falcata, M. phrygia) that were better able to survive than the vetch. These promising species may add to the robustness and versatility of winter-phase legumes for cropping systems in the DPRK.

Year 2005 also saw the second season's data on the effect on rainfall run-off and soil erosion of changing the method of upland maize cultivation to a system that retains crop stubbles (legume and cereal) to protect the soil, and which minimises soil cultivation during crop establishment. The stubble system reduced run-off by 50-70% and soil loss by 60-75% as compared with the traditional maize cultivation practice that has led to severely eroded hill soil and contributed to progression of soil infertility. Wide adoption of this practice will help to preserve the benefits that adopting legumes into cropping systems will bring.

In Australia, progress continues on achieving a better understanding of the inter-relationship between annual legume crops and following cereal crops. Data were collected to help develop a model for simulating N fixation. This parameter is necessary to understand the impact of legume crops on the flow of N between legume, soil and the following cereal. Checking of the CERES wheat model for its accuracy in predicting wheat growth and yield in response to legume crops is well advanced. When accurately modelled, the study will allow exploration of various legume-cereal scenarios to identify opportunities for improving the productivity and sustainability of legume-based cropping systems.