The System of Rice Intensification (SRI) was first evaluated in Tamil Nadu during 2001-2002 through carefully conducted field experiments. Based on the encouraging results from increased yield and from decreased water use, labour for weeding, and seed rate, the State Government accepted a policy proposal to promote the modified method of rice cultivation in Tamil Nadu. The Government sanctioned US$50,000 to evaluate SRI in two major rice-growing areas of the state, one being the Tamiraparani basin in south Tamil Nadu.

We evaluated SRI through 100 Adaptive Research Trials (ARTs) in selected farmers’ fields in different parts of the basin during the wet season of 2003-04. The farmers were exposed to SRI through field demonstrations and theoretical explanations. The ART consisted of comparing SRI with conventional cultivation in 1000 m-2 area without replication. The recommended SRI components were (i) 14-15 d old seedlings from special and simple nursery beds, (ii) planting single seedling per hill spaced at 20 x 20 cm to enable (iii) use of rotary weeder at 10 d interval up to 40-45 days after planting, (iv) irrigating to 2.5 cm depth (up to panicle initiation: after small cracks develop on the soil surface; after panicle initiation: after the ponded water disappeared). Seed bed preparation and planting were done under the supervision of research staff. Rotary weeder was supplied to the farmers. The trials were continuously monitored.

The adoption of all the SRI components depended on local constraints. Only 36 farmers were able to follow all the components, as the others missed some of the components. All the farmers used 14 d old seedlings and adopted 20 x 20 cm spacing. Ten different varieties were used.
Grain yields were recorded carefully by collecting all the panicles from 5 randomly selected 1m-2 areas from both SRI and conventional plots and recording the grain weight after threshing and cleaning. The yield was reported at 14% moisture.
The overall grain yield under SRI and conventional cultivation ranged from 4214 kg ha-1 to 10655 kg ha-1, and from 3887 kg ha-1 to 8730 kg ha-1, respectively. The mean grain yield was 7227 kg ha-1 and 5657 kg ha-1, respectively, showing an overall yield advantage of 1570 kg ha-1 for SRI methods, even used incompletely. Nearly 31 farmers recorded grain yields of more than 8 t ha-1 under SRI, and in the conventional cultivation there were only 3 records of more than 8 t ha-1 yield. The maximum yield advantage recorded for SRI was 4036 kg ha-1 (70%). The yield increase was due to increased number of panicles m-2 and increased number of grains panicle-1. Of the 10 varieties used by the farmers ,three were found to perform very well under SRI. One of these varieties was known for shy tillering with conventional practice, but produced more tillers under SRI.
The benefits of SRI realized by farmers were (i) drastic reduction in seed rate, (ii) no requirement of herbicide, (iii) multiple advantage from using weeder (weed control; less labour for weeding; incorporation top-dressed fertilizer; disturbance and aeration of the soil; increased tillering), (iv) water saving, and (v) increase in number of panicles m-2, grains panicle-1, and grain and straw yield.