Use of genetically modified crops and pesticides in Brazil: growing hazards

1 Departamento de Pesquisa e Desenvolvimento, Empresa Brasileira de Pesquisa Agropecuária. Rodovia Brasília/Anápolis BR 060 Km 09, Gama. 70359-970 Brasília DF Brasil. vicentalmeida@gmail.com 2 Escola Nacional de Saúde Pública, Fiocruz. Rio de Janeiro RJ Brasil. 3 Campanha Permanente Contra os Agrotóxicos e Pela Vida. São Paulo SP Brasil. 4 Associação Brasileira de Agroecologia. Porto Alegre RS Brasil. 5 Fiocruz Ceará. Fortaleza CE Brasil. Use of genetically modified crops and pesticides in Brazil: growing hazards


Introduction
Brazil's status as one of the largest producers of agricultural commodities in the world is associated with an increase in the consumption of agricultural inputs such as pesticides; the national pesticide market was worth US$12.2 billion in 2014 1 .Between 2000 and 2012, the use of these chemicals by unit area more than doubled 2 .This is worrisome because the impacts on environmental and human health due to pesticides have been extensively documented both by international organizations and in the scientific literature [3][4][5][6][7] .
Several studies have pointed to a direct association between the increase in global consumption of pesticides and the use of herbicide-resistant genetically modified (GM) crops 8,9 .In the US, Benbrook 10 revealed that between 1996 and 2011, GM crops led to a 183,000-ton increase in pesticides, which is equivalent to 7% of the overall pesticide use for all crops.Between 1995 and 2002, the use of the herbicide glyphosate in soybean production increased from 2,500 to 30,000 tons/year 8 .During the process of authorization for GM crops resistant to the herbicide 2,4-D, a 3-to 7-fold increase in consumption of 2,4-D was estimated 11 .
In Brazil, GM crops were initially introduced illegally at the end of the 1990s and officially authorized in 2003 12 .Six types of GM crops are authorized, but only three are effectively in use, namely, soybean, corn and cotton.Although genetic manipulation has broader goals such as pharmaceutical applications and biofortified food development, there are currently three types of GM crops currently in use in Brazil: herbicide-resistant, insect-resistant or both 13,14 .In 2014, when pesticide sales in Brazil were the highest, the cultivated area of GM crops reached 42.2 million hectares, which represented an increase of 1306.67% over the 3 million hectares registered in 2003 15 .
In this context, this paper aims to identify and characterize changes in the patterns of use of pesticides and herbicides after the adoption of GM cropsin Brazil.The emphasis is on soybeans, the main commodity produced in the country, in which 90% of the crops are GM according to the ISAAA 15 .The period analyzed was 2000 to 2012, which corresponds to the most recent statistical data on pesticide consumption provided by the Brazilian Institute of Geography and Statistics (IBGE) 16 from the Sustainable Development Indicators publication.This time span also covers the period before and after the official adoption of GM soybean, corn and cotton, which facilitated the analysis of the impactof the adoption of GM crops on pesticide demand.

Methods
In this study, a descriptive research was adopted that was focused on documentary research and based on secondary data under the framework of critical epidemiology 17,18 .This work was developed through systematizing, tabulating, and statistically treating agronomic and demographic data from the IBGE and the Brazilian Crop Protection Industry Union (Sindiveg).
The first step was to calculate for the study period the cumulative growth (∆) and the compound annual growth rate (CAGR) of the following indicators: general pesticide use, pesticide use per capita, pesticide and herbicide use per area, pesticide and herbicide use per cropland, productivity per hectare and population growth.∆ and CAGR were calculated as follows: (1) where V f and V i represent the final and initial values of the analyzed period, respectively, and T represents the difference in years between the final and initial values.
The next step was to carry out a linear correlation analysis between annual pesticide and herbicide use per area and the productivity of each GM crop (soybean, corn and cotton) from 2000 to 2012.The Pearson's correlation coefficient (r) was used to determine the correlation between pesticide use per area (independent variable) and productivity (dependent variable).The determination coefficient (R 2 ) was used to determine the proportion of the variation in productivity that could be predicted from the pesticide use per area.
More specific analyses focused on the changes in patterns of herbicide use and productivity gains for soybean.Genetically modified, herbicide-resistant soybean was the first GM crop officially introduced in Brazil in 2003.

Results and Discussion
During the period investigated, the gross (t) formulated pesticide use in Brazil increased more than 2-fold.Table 1 shows that the cumulative growth (∆) in pesticide use was three times high-er than the growth in productivity (kg/ha) and 10 times higher than population growth for the same period.Each year, pesticide use per capita increased by 7%, while productivity increased by only 3.5%.
Table 2 presents the data for soybean, corn and cotton in Brazil for the analyzed period.Soybean production was associated with a greater than 3-fold increase in pesticide use over the period analyzed (Table 2), while overall pesticide use increased 1.6-fold (Table 1).Furthermore, the determination coefficient (R 2 ) of pesticide use and soybean productivity was 22.73% (Table 2) and of herbicide use and soybean productivity was 17.82% (Figure 1).Figure 2 shows the sudden increase in herbicide use in soybean crops in the year 2003when GM soybean was authorized in Brazil.Herbicide use in corn and cotton also increased, but the change was not as pronounced.
Figure 3 highlights the herbicide use per area (kg/ha) and production of soybean (kg) based on herbicide (kg).Between 2000 and 2002, herbicide use per area decreased by 9% and soybean production per kg herbicide used increased by 18%.From 2003 and onward, herbicide use per area increased by 64% while soybean production per kg herbicide used decreased by 43%.For each ton of herbicide used on soybean crops, there was an annual reduction of 16.79 tons in soybean production (Figure 3).
The cumulative growth of pesticides use in Brazil was higher than the overall productivity of crops between 2000 and 2012.Our data show an increase of 3.2 pp (percentage points) in pesticide use and of 1.78 pp in pesticide use per area, but only 1 pp of productivity increase in the same ,where V f and V i represents the final and initial values of the analyzed period, and T, the difference of years between V f and V i .Share in pesticide use is the average of percentage of pesticide use per crop related to the total amount of pesticide use in the period studied (2000-2012).Productivity represents the amount of crop production per hectare.∆ = (V f / V i ) -1; and CAGR = (V f / V i ) 1/T -1, where V f and V i the final and initial values of the analyzed period, and T, the difference of years between V f and V i .Pearson's correlation coefficient (r) and the determination coefficient (R 2 ) were used in order to achieve the relation between productivity and use of pesticides per area.period (Table 1).This observed increase in pesticide use was not accompanied by an increase in the cultivated area or an increase in the growth of the Brazilian population.These findings disagree with other studies 15,19 that predicted reduced pesticide use after the adoption of GM crops.
A more detailed analysis of the indicator pesticide use per crop showed that only three crops, Herbicides per area (kg/ha) soybeans, corn, and cotton, accounted for 65% of the overall total pesticide use whereas soybean, which is the dominant GM crop, accounted for 71%.The results also show that soybean presented the highest increase in pesticide use per cultivated area and the lowest gain in productivity (Table 2).The indicator pesticide use per area showed that a 1 pp increase in soybean productivity required a 13-pp increase in pesticide use, while for corn and cotton this relationship was approximately 1:1 (Table 2).These data suggest that the genetic modification ofsoybeanwas not associated with growth in productivity and instead contributed to an increase in pesticide use.One explanation for these results is that most GM crops were not developed to enhance productivity or edaphoclimatic adaptability, but mainly to be resistant to herbicides.Other studies have demonstrated that changes in patterns of herbicide use such as an increase in the total amount of glyphosate applied (kg/ha) were related to the adoption of GM soybean 14,20,21 .A study performed in the US between 1990 and 2002 also showed an increase in glyphosate use when herbicide-resistant GM soybeans were authorized (1996) 8 .These changes were not observed for corn and cotton crops, which is likely attributable to the fact that the GM versions of these crops were commercialized in the US at the end of the period of analysis, as also shown in our study.
Some alternative approaches such as increasing the use of different herbicides and development of GM crops resistant to other herbicides have been considered 31,32 .However, these alternatives are also concerning due to the increased probability of serious toxic hazards to humans and the environment by mixing different herbicides 33,34 .It is noteworthy that the two most widely used herbicides in Brazil 2 , glyphosate and 2,4-D,were recently classified as probable and possible carcinogens, respectively, by the International Agency for Research on Cancer (IARC).
The results obtained in this study agree with similar studies in the US, Argentina, and other parts of the world 9,10,35,36 .Data from these studies strongly suggest that the adoption of GM crops increased pesticide use, specifically herbicides sprayed on soybean, as shown by the present study for Brazil.Ecological studies performed in Brazil have shown correlations between soy-bean cultivation (in tons) and mortality due to prostate cancer and between pesticide use and endocrine disturbances 37,38 .As discussed in other studies, data for pesticide and GM crop use may be used as indicators of human and environmental exposure to serious threats, and these data should motivate public actions to prevent or mitigate these hazards 39 .
The results from this study suggest that GM crops have contributed to an increase in pesticide use in Brazil and consequently, increased human and environmental exposure to these potentially hazardous chemical substances.Therefore, the potential for an increase in pesticide use should also be considered during the process of licensing for GM crops.Pesticide use in soybean production increased over the analyzed period, especially after the adoption of GM seeds in 2003.Pesticide use per area also increased significantly, indicating a possible chemical dependency of these croplands and excluding the hypothesis that GM crops could reduce pesticide use.Another relevant aspect, specifically for soybeans, is that this increase did not result in an increase in average productivity.It is also noteworthy that data regarding pesticide use may serve as indicators to support environmental and health surveillance measures such as soil, water and food monitoring for pesticide residues, and such data may strengthen actions related to the diagnosis and treatment of intoxication.

Collaborations
VES Almeida worked on the conception and design of the study, and VES Almeida, K Friedrich, AF Tygel, L Melgarejo and FF Carneiro worked on analysis and data interpretation and manuscript writing and revision.

Table 1 .
Accumulated growth (∆) and compound annual growing rate (CAGR) in pesticide use, agricultural productivity and population between 2000 and 2012 in Brazil.

Table 2 .
Increase of pesticides per crop and area and productivity of genetic modified (GM) crops between 2000 and 2012 in Brazil.