India first allowed the use of GM cotton seeds in 2002. Only ten years later, almost the country’s entire crop is grown using genetically engineered seed. This remarkably fast transition was driven by small farmers deciding that GM seed would improve profitability and reduce insecticide use. Scientists and agronomists initially agreed, producing evidence that the insertion of a natural insecticide (Bt or Bacillus Thuringiensis) into the genes of the plant was the best way of improving India’s historically low cotton yields per hectare. But the last few years have seen optimism fade rapidly as yields have stabilised or fallen and insect resistance has increased. An Indian anti-GM pressure group produced research this week showing that Bt cotton productivity now appears to be falling. (1) As global population increases to about 10 billion in 2050, the world must find ways of increasing the productivity of the limited reserves of usable cropland. Little land is available for conversion from other uses so yields per cropped hectare must grow at close to the rate of population increase. In the past this has proved possible, partly as a result of improved agronomic techniques and hybrid seeds and partly from greater irrigation. Does genetic modification offer a means of continuing the increase as fresh water supplies become stretched? The evidence has been mixed across the world but the Indian experience with cotton is a powerful indication of the issues that can result from GM introduction.

Cotton cultivation in many countries requires huge inputs of pesticide to counter the threat of multiple pests that can reduce yields to virtually nothing. Monsanto’s GM cotton contains one or more genes that produce large concentrations of the natural Bt insecticide in the plant’s leaves. The purpose of the genetic change is to reduce the need for the farmer to spray expensive insecticides which can also severely affect human health.

India has often been touted as strong evidence for the success of Bt cotton, perhaps the country’s most important cash crop. The chart below shows why. Until the turn of the millennium, yields of cotton lint had stagnated at around 300 kilogrammes per hectare of cultivated land. Bt cotton was first officially planted in 2002, though black market seeds were probably in the soil a year earlier. National cotton yields then climbed sharply to levels well over 50% higher. At first sight, the coincident increase in GM plantings and yield increases seems strong evidence for the success of GM.

(Source: Cotton Advisory Board of India for yield figures, SAGE for percentage of GM plantings)

The Indian NGO group, Southern Action on Genetic Engineering (SAGE) points to the possible error in this conclusion. The large part of the yield jump occurred in the first two years after GM introduction. But by that stage only 6% of the cotton planted was Monsanto’s Bt variety. It couldn’t have been the introduction on GM on little more than one twentieth of the land that caused the national increase. Other factors must have played an important role.

The peak year for production per hectare was 2007/08 when yields hit 554 kg per hectare. At this time, 62% of plantings were GM. Since then, the yield has fallen in most years, and is forecast to be 481 kg per hectare in the period to September 2012. SAGE points out that although almost all cotton land in India is now GM, the average yield per hectare will be about the same this year as in 2007/08, when only 6% was planted with GM.

They conclude that GM isn’t helping cotton yields and they are now not alone in their argument. Other NGOs have joined in, railing at the government for encouraging the adaption of Bt cotton a decade ago. But despite the stagnant yields has GM helped in other ways, such as by decreasing the cost of insecticides? The SAGE report says that farmers are now spending 50% more on their agricultural inputs. The seed is more expensive and pesticide use has risen.

So what did cause the sharp rise in yields in the early part of the last decade if it was not the use of GM seeds? One candidate is the increased use of irrigation in Gujarat state. In 2001/02, Gujarat produced 20% of Indian cotton at a yield of 327 kg per hectare, barely above the national average. By 2011/12 projections are for Gujarat yields to be 660 kg per hectare, with the state accounting for 33% of national output. Irrigation seems to have had more impact than GM.

SAGE and other groups have identified several reasons for the apparent failure of GM cotton. First, the insects targeted by the Bt genes have already developed resistance in some parts of India. Other GM crops tagged with Bt genes, such as maize, have begun to see similar problems and so the adaptability of cotton pests should not be a surprise. Second, other pests have moved in to take over. Indian agronomists report increasing problems with pink bollworm, jassids and leaf curl. (As one commentator pointed out ‘in a contest between Monsanto and Darwin, Darwin will always win). Third, GM may have induced a short period of increased yield but this came at the price of decreasing fertility as soil nutrients were drained by the faster growth. To remedy the deficiency farmers will need to increase the use of artificial fertilisers in the future.

We cannot rule out GM on the basis of a poor history for one crop in one country. But the evidence that GM can sustainably increase agricultural yields is still strikingly inconclusive.

(This is part of Chris Goodall’s forthcoming book, Sustainability: All That Matters, to be published by Hodder later this year).

[1] http://www.ddsindia.com/www/pdf/PR%20April%209,%202012.pdf