Herbicide Resistance in Brazil: Status, Impacts, and Future Challenges

3.1 Resistance to ALS inhibitors

The first cases of resistance to ALS inhibitors were Euphorbia heterophylla and Bidens pilosa reported in 1993 in soybean areas in the states of Mato Grosso do Sul and Rio Grande do Sul, which showed cross-resistance to sulfonylureas and imidazolinones [26]. After, resistant biotypes of B. subalternans (1996) [27], Parthenium hysterophorus (2004) [28], Conyza sumatrensis (2011) [29] and Ageratum conyzoides (2013) were found in Paraná. The latter species was also reported in cotton in Mato Grosso [11]. However, the greatest resistance challenges to ALS inhibitors are found in irrigated rice cultivation. The species reported with ALS resistance in this culture are: Sagittaria montevidensis (1999) [30], Echinochloa sp. (1999) [31], Cyperus difformis (2000) [32], Fimbristylis miliaceae (2001), Oryza sativa (2006), and Cyperus iria (2014) [33] in Rio Grande do Sul and Santa Catarina.

Cases of resistance in rice cultivation are associated with the rapid adoption of Clearfield® technology (crops tolerant to imidazolinones, a chemical group of ALS inhibitors), which were introduced in 2002 in areas of southern Brazil [34]. Although the emergence of new resistant species after the adoption of Clearfield® cultivars did not increase significantly, the dispersion of weed populations resistant to ALS inhibitors, mainly of red rice, was favored by genetic flow of cultivated rice to red rice, representing a great agricultural, economic, and social restriction in the use of Clearfield® technology [35].

Other specific, but not least, cases of resistance to ALS inhibitors are Raphanus sativus (2001), Lolium multiflorum (2010), and R. raphanistrum (2013), found in wheat and barley in Rio Grande do Sul and Paraná; and Amaranthus retroflexus (2012) in cotton in the states of Mato Grosso, Mato Grosso do Sul and Goiás [11, 36, 37].

3.2 Resistance to EPSPs inhibitors

Currently, nine weed species have been reported with glyphosate resistance in Brazil, some of these species have multiple resistance to other modes of action [11]. Lolium multiflorum (2003) was the first species identified with glyphosate resistance in orchards and vineyards from Rio Grande do Sul [38]. After, Conyza bonariensis (2005), C. canadensis (2005) [39], C. sumatrensis (2010) [40], Digitaria insularis (2008) [12], Chloris elata (2014) [13], Amaranthus palmeri (2015) [14], Eleusine indica [15], and A. hybridus (2018) [11] were identified with this resistance mainly in maize and soybean, and wheat fields, but also in citrus and coffee orchards in the states of Mato Grosso, Paraná, Rio Grande do Sul and São Paulo.

With the exception of L. multiflorum, the selection of glyphosate resistance in these species is related to the use of GR cultivars, which has also influenced their dispersion throughout the country. Resistant populations of L. multiflorm have gone from infesting apple orchards and vineyards to invading GR-soybean fields in the southern states of Brazil [41]. The species of the genus Conyza, which have a high invasive potential due to the large seed production, the rapid and high germination capacity, cause great damage to agriculture, and due to their poor interspecific differentiation, it can be an exchange of resistant alleles between species [42]. However, D. insularis has been, among glyphosate resistant species, one of the main problems to be faced; therefore, greater efforts have been made to characterize the factors involved in its resistance, dispersal and management [12, 43, 44, 45, 46, 47, 48]. Molecular studies showed that the first glyphosate resistant D. insularis populations found in the country (Guairá—Paraná) came from Paraguay and were dispersed to other states of Brazil, partly due to their biology and perennial capacity, but mainly due to anthropogenic activities, such as the lack of cleanliness of agricultural implements, but also events of independent selection [47, 48].

Other weeds that pose a major challenge to Brazilian agriculture are species of the genus Amaranthus, as they are often reported with glyphosate resistance in GR fields in the United States and Argentina [49]. In addition, Amaranthus sp. can hybridize interspecifically facilitating dispersion of resistance alleles [50]. In Brazil, A. palmeri was reported to have glyphosate resistance in 2015 [11], when its multiple resistance to the ALS inhibitors was also corroborated [14]. However, the Instituto Mato-Grossense do Algodão had records of the occurrence of glyphosate resistant populations of this species since 2012 in the municipalities of Ipiranga do Norte and Tepurah, Mato Grosso, that was imported from Argentina in cotton harvesting machines in 2011 [51]. Recently, multiple resistance of A. hybridus to glyphosate and ALS inhibitors was also confirmed in Rio Grande do Sul in soybeans [11, 52]. With respect to the latter case, there is great concern because it is feared that it has also been introduced from Argentina, where populations of A. hybridus with this resistance profile carry mutations in the genes encoding the target enzymes [53]. In the case of glyphosate resistance, it is a triple mutation that confers high levels of resistance and that had not previously been observed in any other species [10, 54]. In addition, in Argentina there are also populations of the species with multiple resistance to 2,4-D and dicamba [55]. Therefore, if it is confirmed that the resistant populations of A. hybridus found in Brazil were introduced from Argentina, the scenario faced by Brazilian farmers in the coming years in relation to weed management will be very difficult.

3.3 Resistance to ACCase inhibitors

Urochloa plantaginea (1997) [56], Digitaria ciliaris (2002) [57], Eleusine indica (2003) [58], Avena fatua (2010) [11], and D. insularis [16] were reported with resistance to ACCase inhibitors, mainly in non-transgenic soybean fields. These findings demonstrate the importance of these herbicides for the control of grasses in soybean fields, due to the low availability of selective herbicides that effectively control these weeds in pre-emergence conditions, allied to the difficulties of using graminicides, since these products have high retention in the organic matter [29].

3.4 Resistance to other mechanisms of action

The majority of herbicide resistance cases reported in Brazil are included in the three groups of herbicides described above, following the global trend. However, cases of resistance to other modes of action have also been found. In 1999, Echinochloa crus-pavonis and E. crus-galli were reported with resistance to synthetic auxins, specifically quinclorac, in rice fields of Itajai, Santa Catarina [59]. Amaranthus retroflexus (2014) and C. sumatrensis (2017) were reported with resistance to PPO inhibitors [11]. The first showed fomesafen resistance and it was found in GR-soybean and -cotton fields of Mato Grosso; and C. sumatrensis presented resistance to saflufenacil in soybean fields in the western region of Paraná in the municipalities of Palotina and Assis Chateaubriand [11]. This last species had already been confirmed to be resistance to chlorimuron-ethyl (ALS inhibitor) in 2011 [60] and paraquat (PSI inhibitor) in 2016 [61] within the same region.

3.5 Cross- and multiple-resistance

Cross resistance is expressed when a weed resistant biotype shown resistance against two or more herbicides with the same mode of action, and multiple resistance occurs when a weed resistant to a given herbicide manifests resistance to two or more different modes of action. Most cases of resistance to ALS inhibitors have cross resistance, that is, weeds resistant to imidazolinones often have a degree of resistance to sulfunylureas and vice versa [29]. Eleusine indica resistant to sethoxydim (cyclohexanediones) showed resistance to the ariloxifenoxipropionatos (FOPs) [58], and quinclorac resistant E. crus-galli showed cross resistant to others synthetic auxins [11]. Weeds with cross resistance represent a great challenge for Brazilian agricultural sustainability; however, weeds with multiple resistance are more challenging by reducing chemical alternatives for their control.

The occurrence of multiple resistance has increased significantly in recent years, and most of the reported cases occurred in the post-glyphosate era. The first case of multiple resistance was E. heterophylla, which was found in fields of maize and soybeans in 2004 and showed resistant to triclopyr and fomesafen (ALS + PPO) [62]. In 2009, E. crus-galli was found with resistance to synthetic auxins and ALS inhibitors in rice fields in Rio Grande do Sul [34]. Biotypes of B. subalternans (2006) and B. pilosa (2016) were found to be resistant to atrazine (PSII inhibitors) and ALS inhibitors in soybean and maize fields from Paraná [63]. Among the cases that involves glyphosate resistance are C. sumatrensis (2014), A. palmeri (2015) and A. hybridus (2018) as dicots, that also shown resistance to the ALS inhibitors and were found in soybean fields [11, 14, 60], and L. multiflorum (2010), D. insularis (2016), and E. indica (2016) as monocots with resistance to the ACCase inhibitors. However, the most worrying case is Conyza sumatrensis reported in 2017, which was found as being resistant to EPSPs, PSI, PSII, PPO and synthetic auxins in a GR-soybean field from Assis Chateaubriand-PR [11].

This brief account shows the global scenario of the current situation of herbicide resistance in Brazil; however, it is far from reality, because only the first occurrence of a unique case (species x site of action) is reported, while in countries like the United States and Australia, there are multiple reports for the same unique case of herbicide resistant occurring in different regions. For example, the case of A. palmeri resistant to glyphosate have more than 30 reports along of the United States [11]. To have an idea of the real problem in Brazil, we have as an example the study conducted by Lopez-Ovejero et al. [45], who determined the frequency and dispersion patterns of glyphosate resistant D. insularis revealing the existence of 1299 (of 2596) populations with different resistance levels to this herbicide distributed only in the areas of soybean production. In the scientific-academic environment it is commonly said that it is more difficult to find a population susceptible to the glyphosate of C. sumatrensis or D. insularis than a resistant one. In addition, from the botanical point of view, more species of the Amaranthaceae, Asteraceae, Cyperaceae, and Poaceae families have high potential to select for resistant to the inhibitors of ALS, ACCase, EPSPs, PPO, and synthetic auxin herbicides in the coming years [64].

4.1 Agronomic factors

Among the agronomic factors that favored the rapid selection of resistance are the characteristics of the herbicide used and the cultural practices. Some herbicide chemical groups have a higher risk of selecting for resistance, especially those with a single mechanism of action or detoxification way (high specificity). High dose applications provide greater selection pressure for resistant weed individuals. The greater persistence of a herbicide also favors the selection for resistance, since the period of exposure is longer, therefore, the ideal is that the herbicide only has effect in the critical period of competence. Reduced crop rotation (monoculture), lack of alternative herbicides, nonuse of herbicide mixtures or sequential applications, nonremoval of weeds from field that escaped herbicide control, and poor inclusion of nonchemical methods are major cultural practices that can lead to emergence of herbicide resistance [66].

In Brazil, a large part of crop production systems is intensive, and today effective weed management without herbicides is inconceivable in the short term in these systems [67]. In addition, a large part of the agricultural areas is occupied with HR crops, resistant to glyphosate or imidazolines, as described in Section 2. The adoption and the use of these technologies caused great changes in weed management, which in most cases, implied the substitution of different herbicidal molecules, that were traditionally used before the insertion of HR crops, by the almost exclusive herbicide associated with said technology in question, at least in the first years after its adoption [23, 68]. For example, in the United States, glyphosate applications replaced a large part of previously used herbicides in GR crops [69].

In Brazil, during the first years after the adoption of GR crops, glyphosate was used in various steps of the production process (chemical fallow (pre-planting), weed management (single or sequential), and desiccation) in doses ranging from 2 to 8 L ha⁻¹, and in some cases, those doses exceeded 10 L ha⁻¹ per application [69]. In other cases, many GR soybean farmers delayed the management of weeds that germinated before planting in order to control them with post-emergent applications of glyphosate made on the crop when the competition between the soybean and weeds had already begun [23, 69].

The almost exclusive use of glyphosate quickly showed deficiencies in weed control [23]. Species such as A. palmeri, Conyza sp., C. elata, D. insularis, and E. indica selected for resistance to this herbicide, forcing farmers to use other herbicides in areas cultivated with GR crops [70]. Herbicides such as 2,4-D, ACCase inhibitors, and ALS were retaken for weed control during pre-sowing (chemical fallow) and crop development, and glufosinate, diuron, and paraquat for desiccation. Currently, glyphosate is applied in isolation only 14% of the time [71]. At the same time, the relative amount of glyphosate used per hectare decreased. For example, 118.5 tons of glyphosate were sold in 2009 and there were 18.6 million ha of GR crops (14.1 soybean +4.4 maize +0.13 cotton), and by 2018, there were 49.2 million ha of GR crops (33.4 soybean +14.7 maize +1.15 cotton) and 195.1 tons of glyphosate were sold. Considering that only these three crops consume 70% of pesticides market of Brazil, in 2009, 4.46 kg of glyphosate ha⁻¹ year⁻¹ were used, while in 2018, that amount was 2.78 kg of glyphosate ha⁻¹ year⁻¹, i.e., there was a reduction of at least 26% (Figure 1A and 2A). On the other hand, sales of herbicides such as 2,4-D, paraquat, atrazine, increased between 2009 and 2018 (Figure 2A). However, the increase in the use of herbicides with different mode of action, applied in mixture or in sequence with glyphosate, has contributed to the emergence of weeds with multiple resistance.

Paraquat and diuron are considered as bodyguard of glyphosate and are essential tools for Brazilian farmers to hamper the spread of glyphosate resistant weeds [72]. However, the use of paraquat is only authorized until 2020 by the Agência Nacional de Vigilância Sanitaria (ANVISA) after several studies demonstrated that this herbicide can cause Parkinson’s and irreversible damage to the genome [73]. Therefore, this legal determination will represent a new challenge in relation to the management of glyphosate resistant weeds, not only in GR fields.

Currently, soybean farmers are anxious and have high expectations with the introduction of new varieties of transgenic soybeans of the technologies Enlist E3™ (2,4-D + glyphosate + glufosinate) and Intacta 2 Xtend® (dicamba + glyphosate) that, have stacked traits of resistance to lepidoptera and up to three herbicides and, will be available in the Brazilian market for commercial use as of the 2020/2021 and 2021/2022 cycles, respectively [74, 75]. On the other hand, the use of synthetic auxins has also been questioned. According to the Instituto Brasileiro do Vinho (Ibravin), the 2,4-D drift used to control weeds in pre-planting of GR soybeans caused damage estimated in R$ 100 million only in 2018 in vineyards of Rio Grande do Sul [76]. Moreover, the use of auxinic herbicides needs to be done with caution, especially in periods with lower temperatures, since any problems related to the application technology, which allows the contact of Digitaria sp. plants with low doses of these herbicides, will promote the re-growth of these weeds, which will hinder its control and favor the dispersion of this species [77].

Crop rotation is a consolidated weed management strategy in most of the grain-producing agricultural regions of Brazil. However, it often involves the use of the same technology, i.e., GR soybean is replaced by GR maize and vice versa. This situation is due to the high competitiveness of global commodity markets, which have led farmers to specialize in the production of one or few closely related crops, avoiding the implementation of more complex crop rotations (grain by vegetables). The efficiency obtained by the specialization, which allows the use of the same seeder, combine and marketing infrastructure, has led to the widespread adoption of monocultures [67]. This limited crop rotation (grains by grains), has impacted on the use of herbicides because the number of applications is doubled per agricultural year (3–5 applications per agricultural summer or winter cycle), since second crop requires similar agricultural tasks to the first crop. This practice has increased the herbicide selection pressure on weed populations, but also have provoked the occurrence of voluntary plants from the previous crop, which are difficult to control because they have a similar herbicide resistance profile as the current crop, reducing the crop yield [78, 79].

Direct sowing systems in conjunction with other cultural weed control practices such as cover crops and crop rotations reduce weed population densities [67]. Brazil is one of the few countries that have widely adopted the direct sowing system. This production system reduce the annual weed density compared to conventional agricultural systems [67]; however, the appearance of biannual (Conyza sp.) or perennial (C. elata, D. insularis, and E. indica) weeds have been favored [29, 40, 66], which under continuous selection pressure of glyphosate selected for resistance; therefore, weed management strategies more complex are required.

4.2 Scientific and technical aspects

The Brazilian technical and scientific community specialized on weed science is very large, and their efforts to prevent, monitor, identify, and establish management programs of herbicide resistant weeds, as well as to alert farmers about the occurrence of new cases in order to reduce their dispersion are also very large. Symposiums, congresses, and multiple regional, national, and even international extension activities are frequently held to improve the sustainability of the main production systems, bringing together farmers, researchers, agricultural companies (machinery, pesticides, seeds, etc.), politicians, and agronomy students. In this way we can affirm that the Brazilian agricultural community knows in depth the negative impacts of herbicide resistance. The main efforts of the scientific community are addressed in combating the “problem,” as evidenced by the greater amount of research papers proposing alternative management strategies of herbicide resistant weeds. For example, in 2019, of the 150 articles published in Planta Daninha (scientific journal published by the Brazilian Society of Weed Science), 22 papers addressed issues related to alternative methods of weed management (chemical, nonchemical, or combined measures), four articles reported the occurrence of new herbicide resistant cases, and only two articles fully or partially characterized the mechanisms of resistance involved (Material S1), i.e., studied the “cause” of herbicide resistance. Planta Daninha is not the only scientific journal where Brazilian researchers publish their results, but if it is the main one; therefore, these data reflect the trend in which they invest their main efforts to combat herbicide resistance.

Knowing the “problem” is one thing, but knowing the “cause” is another. Herbicide resistance would be equivalent to referring to a headache. In both cases, the “problem” is known, but the “cause” is unknown. We often underestimate headaches (which can be caused by muscle tension, stress, anxiety, head trauma, etc.) by resorting to self-medication or requesting medication from the pharmacist on duty, who asks a series of questions and recommends some type of analgesic. The pharmacist has not identified the “cause” of the problem, but his recommendation could totally or temporarily relieve the headache and, at the same time, we avoid the consultation with a specialist doctor. Similarly, agricultural field technicians have a deep understanding of the negative impact of herbicide resistant weeds and often recommend different management alternatives; however, they do not know the “cause” of herbicide resistance. Implementing herbicide resistance management measures without knowing the cause of it, by characterizing the resistance mechanisms that govern it would be equivalent to self-medicating. In many cases, herbicide resistant weeds are satisfactory controlled initially, but often the problem worsens over time, resulting in cases of cross and/or multiple resistance. This analogy allows us to infer, that in many cases, Brazilian technicians, and even weed scientist, have acted more as pharmacists than as doctors. This scenario can be added that many field professionals (agronomists and sales agents of pesticides) act without professional ethics prescribing pesticides in a superhuman rhythm [80]. In Brazil, farmers need of prescriptions to purchase these products; however, a professional is often an employer of cooperatives or reseller pesticide offices, so he needs to sell supplies to guarantee his employment [80]. A study carried out by the Agência de Defesa Agropecuária do Paraná (ADAPAR) showed that at least 30% (600 of 2000) of the field professionals signed daily between 7 and 17 prescriptions between 2015 and 2017, i.e., they signed 1–2 prescription for every hour of work. According to ADAPAR, this fact is technically inviable due to the long distance between properties, because to sing a prescription, the field professional must visit the crop fields [80].

In theory, Brazilian weed scientists know the cause of herbicide resistance, i.e., they are familiar with the possible physiological, biochemical, and molecular mechanisms that can confer it. However, studies aimed at characterizing these mechanisms are scarce, often conducted only with the first population(s) that confirmed the occurrence of a given case of herbicide resistance. In addition, the resistance mechanisms characterized in a herbicide resistant population have been adopted in a generalized manner by technicians and other weed researchers, assuming that new occurrences of a case of herbicide resistance already reported (species x herbicide) will have the same mechanisms observed in the first resistant populations. This conclusion may be partially correct when a new occurrence is found in agricultural areas near where the first occurrence was found (dispersion) [48]. However, resistance within the same area and between geographically distant areas may be due to independent herbicide resistance selection events [81], so the resistance mechanisms involved may be different. When multiple studies on the characterization of resistance mechanisms have been carried out on the same weed, the results have been interpreted in a controversial way by the scientific community. For example, in the case of glyphosate resistant D. insularis there is no consensus of the mechanism that governs such resistance [82]. In the first populations of glyphosate-resistant D. insularis, collected in the state of São Paulo in 2009, the reduced absorption, translocation and metabolism of the herbicide, and a gene mutation (at the Pro-106 EPSPs gene position referred to as Pro-182) were the mechanisms conferring that resistance [12]. After, differences in absorption, but not in the translocation of glyphosate and the occurrence of mutations were observed in other populations [83, 84], while collected populations in different regions of the State of São Paulo presented mutations and enhanced activity of the EPSPs [85]. In the most recent study, including glyphosate resistant D. insularis populations collected in different states, it was not possible to characterize the mechanism (s) involved in the resistance [86]. These results show that resistance to a given herbicide can be governed by different mechanisms, acting in isolation or together, in the same species. In addition, these divergent results observed in the different studies show that each new occurrence must be evaluated individually, therefore, this information should not be used to generate a consensus on the mechanisms involved in the resistance of a given case (herbicide x species).

Knowing the mechanisms that govern herbicide resistance is fundamental to plan a proper management strategy, since in some cases, a specific mutation (target site mechanism) does not represent that a herbicide with the same mode of action to which resistance was observed can be used. For example, the Ala-122-Thr mutation in the ALS gene confers high resistance to imidazolinones, but does not confer resistance to sulfonylureas [87]. In the case of herbicide metabolism (non-target-site mechanism) regulated by the cytochrome P-450 enzyme complex, which can confer multiple resistance up to six or possibly more groups of herbicides [88], the use of the same herbicide to which resistance was reported it may be possible using a cytochrome P-450 inhibitor such as malathion or phorate before applying the herbicide in question [89, 90]. Paraquat is an alternative for the management of glyphosate resistant weeds only if resistance is not governed by vacuolar sequestration, since although they have different mechanisms of action, this non-target-site mechanism confers resistance to both herbicides [91]. It is important to note that the management of herbicide resistance is not as simple as described here, since it often involves the participation of different resistance mechanisms, but the timely and appropriate characterization of them could reduce the use of complex mixtures of herbicides in high doses, reducing the impact environmental [92].

Some Brazilian weed research groups have partnerships with weed scientists from the United States and Spain and other countries, who collaborate actively in studies that characterize the mechanisms of the most important cases of herbicide resistance of the country [12, 13, 14, 15, 16, 93]. However, these efforts are still insufficient and the previous information allows us to infer that in most cases, the management of herbicide resistance in Brazil has been faced in an inverse way, i.e., first, efforts are made to test and propose different alternatives to solve the “problem” and, in some cases, then try to characterize the “cause.” This would be equivalent to placing the shoes first and then the socks. Therefore, to face the problem of herbicide resistance, more efforts must be made to characterize the resistance mechanisms involved in each case, and only later, evaluate different alternative management strategies that are technical and economically viable.

4.3 Economic factors

In this aspect there are two great scenarios. On the one hand, there are the pesticide manufacturers that are valued according to the volume of sales of their products; therefore, they invest their efforts in “conquering” more farmers every day so that they use their products and thus have greater presence in the market and consequently greater prestige. On the other hand, there are farmers who in turn want to obtain the highest profit margin with the least investment, often, in the short term.

As highlighted in the previous section, Brazilian scientists focus their efforts on developing herbicide-resistant weed management strategies, mainly through the applications of herbicides with different modes of action applied in tank mix or in sequence [71]. These investigations are often funded by pesticide manufacturers. Although the conclusions are not biased, objectively reflecting which treatments are the best alternative to control certain weed resistant herbicide(s) in a particular production system, and the researchers also do not recommend the use of commercial formulations of a specific manufacturer, obviously the intention of the financing pesticide manufacturer is to increase the sales of its products and technologies.

Weed researchers evaluating alternative management programs often find at least one efficient control option, both for the level of control achieved (> 80%) and for the period that a treatment maintains the level of control, i.e., there are solutions to the “problem,” and Brazilian weed scientists never have stopped looking for new herbicide management alternatives. However, if research is abundant in this regard in the country, why do cases of herbicide resistance continue to increase? The answer to this question is possibly related to the fact that in most of these studies the costs (herbicides + cost of operations + worker’s payment) of the resistance management programs evaluated are not considered. In addition, the yield (kg ha⁻¹) that a given management program can guarantee to the farmer is rarely determined. A specific case that addresses these two aspects (cost vs. yield) is the study developed by Piasecki et al. [94], who evaluated 16 treatments, of which 11 did not show differences in soybean yield (3600–3750 kg ha⁻¹), but there were differences in the costs of each treatment, since they were composed by 3 or 4 herbicides. In that study, the highest yield of soybean (3888 kg ha⁻¹) was achieved with the treatment consisting of glyphosate + chlorimuron-ethyl +2,4-D + saflufenacil (T13: 1080 + 25 + 670 + 50 g ia ha⁻¹), which had a cost of R$ 180.00 ha⁻¹. However, the best relative economic return was obtained with the treatment of glyphosate + chlorimuron-ethyl +2,4-D (T12: 1080 + 22.5 + 670 g ia ha⁻¹), which presented a yield of 3749 kg ha⁻¹ and cost only R$ 85.00 ha⁻¹. This study did not include the costs related to the application operations, but contrasting the cost of a management program with the crop yield can be an additional tool for the farmer, so that he can estimate his profit margin and decide whether or not to adopt given weed management program.

This situation is also reflected in the type of HR crop technology used by farmers. For example, Liberty Link® technology (glufosinate resistant crops) is available in Brazil since 2016/2017 cycle [95]; however, its use is low compared to GR crops, since glufosinate is, in average, three times more expensive than glyphosate. Total glufosinate sales exceeded 1000 tons year⁻¹ in 2017 (1137 tons) and 2018 (1450 tons), but they are still very far from glyphosate sales (173,150 and 195,056 tons in 2017 and 2018, respectively) [7]. This shown that farmers often prefer to continue living with glyphosate resistance than to adopt a new but more expensive technologies, i.e., the adoption of an HR technology is motivated by the cost–benefit ratio by saving costs devoted to pest control guaranteeing high yields [2]. Therefore, the success of Enlist E3™ and Intacta 2 Xtend® technologies, which will be available in the Brazilian market from 2020/2021 and 2021/2022 crop cycles, respectively, will depend on their final cost; meanwhile, farmers will continue to be reluctant to adopt integrated management measures for herbicide-resistant weed control or new HR technologies [96, 97]. The Brazilian scientific community has the task of demonstrating to the farmers that, although the implementation of an integrated weed management program is complex and expensive initially, in the long term it is profitable and environmentally sustainable [98].