A. Introduction

In 2001/02 Argentina grew 152,000 hectares of cotton with an average lint yield of 328 kg/hectare for a total production of 50,000 tons of lint (ICAC 2002a). Cotton area in Argentina fluctuates with the international price of cotton and has declined from around 750,000 hectares in 1998/99 to 150,000 hectares in 2001/02. Bt cotton developed by Monsanto and Delta Pine Land, with the Cry1Ac gene, was successfully field tested in Argentina and approved for commercialization in 1998 (James 1998). Most of the cotton in Argentina is grown in two provinces in the north east, in Chaco and Santiago del Estero, which together grow almost 90% of the crop. Cotton is grown on large farms, greater than 90 hectares, which account for 70% of production, on small farms that account for 21% of production, and on the smallest farms, called minifundios (less than 20 hectares) which account for 9% of cotton production. The total number of cotton farmers in Argentina in 2000/01 was approximately 10,000 (Qaim 2002). Two studies have been conducted to assess the economic impact and to characterize the adoption of Bt cotton and the results are summarized in this case study (Qaim and de Janvry 2002, Elena 2001).

B. Insect Pests of Cotton

The major cotton insect pests in Argentina that can be controlled with Bt are cotton bollworm (Helicoverpa armigera), tobacco budworm (Heliothis virescens), cotton leafworm (Alabama argillacea) and the pink bollworm (Pectinophora gossypiella). The two Bt cotton varieties, incorporating the Cry1Ac Bt gene, NuCOTN 33B and DP 50B have been released in Argentina.

C. Adoption of Bt Cotton

Bt cotton was introduced in 1998 with a small hectarage of 5,500 hectares equivalent to 0.7% of the total 750,930 hectares (Qaim and de Janvry, 2002). The area of Bt cotton increased from 5,500 hectares in 1998 to 12,000 in 1999 (3.6% of cotton area), to 22,000 hectares in 2000 (5.4%) and was static at 5% in 2001/02 (Table 43). These adoption figures are consistent with those reported by James (1999, 2001a).

D. Comparison between the performance of Bt cotton and non-Bt cotton

The data in Table 44 for yield, insecticide usage, cost of seed and overall economic advantage for Bt cotton and non-Bt cotton are averages for the two seasons, 1999/2000 and 2000/2001, derived from the data reported by Qaim and de Janvry (2002). A significant and consistent yield advantage of approximately 35% was generated by Bt cotton compared with non-Bt cotton (Table 44). The results from the survey conducted in 1999/2000 (Elena 2001) also report an increased yield of 907 kg/hectare for Bt cotton.

Non-Bt cotton on average required 4.8 sprays, compared with 2.5 sprays on Bt cotton, i.e. 2.3 fewer insecticide sprays. Compared with conventional cotton, in 1999/00, Bt cotton required 2.1 sprays compared with 4.5 sprays, for a difference of 2.4 fewer sprays, and in 2000/01, 2.8 sprays compared with 5.1 for non-Bt cotton, for a difference of 2.3 fewer sprays (Qaim 2002). On average Bt cotton required half the number of sprays. The data in Table 44 show a decrease of approximately 50% in cost of insecticides on Bt cotton, compared with non-Bt which translated to an economic gain of approximately $17.50/hectare. Elena (2001) also reports a corresponding gain of $27.55 in favor of Bt cotton in 1999/00 due to reduced insecticides.

The cost of Bt cotton seed was $103/hectare compared with $17/hectare for conventional seed; a six fold difference in price equivalent to approximately $85/hectare (Table 44). Data from the survey conducted in the same areas in 1999/00 (Elena 2001) estimated the difference in price between Bt cotton and non-Bt cotton at $75/hectare which is slightly lower, but in the same range.

Gross margin advantage for Bt cotton was estimated at an average of $20/hectare, with considerable variation between 1999/00 and 2000/01.The lower gross margins for Bt in 2000/01, ($5/hectare) compared with 1999/00 ($36/hectare), were influenced by the lower price of cotton in 2000/01 and higher input costs. Elena (2001) reported a gross margin of $65 per hectare in favor of Bt cotton for 1999/2001.

E. Concluding Remarks

Given that the lepidopteran insect pests which lend themselves for control by Bt are important in Argentina and that Bt farmers benefit from increased yield, decreased cost of insecticides and higher net returns, the issue of why adoption of Bt cotton has stagnated at 5% deserves to be addressed (Qaim and de Janvry 2002). There is limited awareness of the benefits of Bt cotton, particularly amongst small farmers (less than 20 hectares). However, Qaim and de Janvry (2002) conclude that the major constraint to adoption is the high price ($103/hectare) of Bt cotton seed, compared with $17/hectare for non-Bt cotton seed. For the average cotton grower they note that the high price of Bt cotton seed can more than double farmer expenditure for all purchased inputs. They conclude that a price of $103 is equivalent to a $78 technology fee, which is similar to the USA, but pest infestation levels in Argentina are generally lower than in the USA, leading to lower returns in a cotton-growing system which is subsidy-free and lower-cost than the USA. For all these reasons, the value of Bt cotton is lower in Argentina than the USA and hence Qaim and de Janvry suggest that it may be appropriate to consider pricing Bt cotton seed at a lower level than the USA, as is done in Mexico (Traxler et al 2001). A simulated demand curve for Bt cotton in Argentina indicated that the optimal price for both farmers (at a 40 to 50% adoption rate) and the developers of the technology would be a price of about $40 to $50/hectare for Bt seed (Qaim and de Janvry 2002).

F. Highlights