Changing rice varieties

It is by now well-documented that the farmers in Bangladesh (and also in India) have replaced the traditional varieties of rice (TVs) with modern ones (MVs). This was triggered mainly by higher yield but other considerations have also played a part in influencing the decision.  Whatever might have been the reasons, the critics of the Green Revolution detected two important drawbacks of such wholesale transformation in adoption. First, a shift from TVs to MVs has caused loss of crop diversity in the wake of a large number of TVs being replaced by a few MVs. Second, farmers tend to grow the same varieties that allegedly impact upon soil fertility. It is unfortunate that attempts to test these two hypotheses are rare, if not absent, among contemporary empirics. It is, however, heartening to note that a book titled, Adoption and diffusion of modern varieties in Bangladesh and eastern India (edited by Mahabub Hossain and others, 2012)   sheds some important light on the issues at stake.  
  A survey of sample households in 62 villages carried out by BRAC in 2014 showed that in rainy Aman season, farmers generally grow 325 varieties of rice; in Boro season, the number is 96. This can be pitted against 157 and 64 varieties respectively in 2000. This shows that the number of varieties went up over time possibly with no impacts on crop diversity. To add more meat to the argument, a comparison of a large-scale survey conducted by the Department of Agricultural Extension (DAE) in 1996 and the IRRI in 2005 showed that the number of varieties in both seasons increased by a big margin although the number heavily declined in the Aus season. The recent research shows that more than 1,000 varieties are grown in the three seasons. More importantly, TVs still dominate some areas because all plots are not equally suited to modern varieties, unfavourable zones and preference for puffed and fine quality rice.
    Let us now look at the concentration of varieties grown. In the Boro season, roughly 60 per cent of the cultivated land is occupied only by two varieties -  Brridhan 28 (about 32 per cent) and Brridhan 29  (28 per cent).  Next comes Heera accounting for 12 per cent. However, in 2000, the combined share of the two varieties was 21 per cent only. This means a many fold increase over time in the share of the two modern varieties. Surveys undertaken in 2004/05 also indicate that these two MVs claimed 55 per cent of sown areas. The surprising thing to note is the huge preference by farmers for Brridhan 28 over Brridhan 29 despite the latter fetching one ton more per hectare than the former (6 tons vs. 5 tons). In this particular scenario, the yield advantage seems to have been outshone by other factors such as quick harvest in 140 days against 160 (and thus providing an advantage for growing other crops), higher price per maund possibly due to better taste, suitability in medium land as compared to Brridhan 29 grown in low land that hardly provides scope for other crops.  Besides, in a few cases, Brridhan 28 is also grown in Aus season with of course relatively low yield than Boro season. The second point to note is that newly released varieties,  such as Brridhans 33,40,39,47,50 and 49 are yet to penetrate on a large scale as most of them are ecology-specific. Extensional weakness could be another reason. It would perhaps be better if the replaced varieties carried earlier names with added attribute - for example, 'Brridhan 28 salinity resistant' or 'Brridhan 29 resistant to water logging'. It is because adoption often assumes cultural phenomena with generations adopting the variety with same name.
Has the yield of MVs declined over time? It is a concern that is aired often. Recent data show no decrease in yield rate as often alleged. It is true that yield rate is the major determinant of adoption of a particular variety. But farmers' choice about variety also depends on shorter maturity and eating quality, such as preference of Brridhan 28 over 29. It also depends on access to information and nearness to market. The main source of first information is the DAE staff although it reaches the rich first. The second source of information is other farmers followed by relatives and friends. Indian varieties are spreading through neighbours and fertiliser dealers while hybrid is spread mostly through dealers. The main source of seed is farmers' own harvest.
Considerable diversity still exists for rice varieties at farm level. The diversity is higher in Aman and Aus seasons than in Boro season.  Many of the popular varieties are on way to extinction. The important extinct Aman varieties are Nazir Shail, Loti Shail, Vaduri etc. The main reasons for parting with popular varieties are low yield, long maturity and logging problems. It is not that only TVs are shown the door; MVs are also being replaced by released varieties. Especially with climate change in the horizon, more research and extension work are needed to develop newer varieties and for this, higher allocation is required for agricultural research and extension in the national budget.
Abdul Bayes is a Professor of Economics at Jahangirnagar University.
[email protected]