**3. Conventional and RR genetically modified soybeans: Some results in Brazil**

#### **3.1 Physiological quality and lignin content in the seed coats submitted to different harvest times**

The viability period of the soybean seed is extremely variable, depending both on genetic characteristics and environmental effects during the phases of development, harvest, processing and storage. Once unfavorable conditions occur in some of these phases, physiological damages may result in losses to seed quality, with the intensity of these damages varying with the genetic factors intrinsic to each cultivar. Various researchers have emphasized the possibility of use of the seed with seed coat with a certain degree of impermeability to water as an alternative for avoiding loss of quality in the field (Gilioli & França Neto, 1982; Peske & Pereira, 1983; Hartwig & Potts, 1987), with delay in harvest and determination of the lignin content in the seed coat being methodologies suggested for genetic breeding programs for evaluation of the quality of soybean seeds (França Neto & Krzyzanowski, 2003).

Within this context, the work presented below (Gris et al., 2010) was conducted with the purpose of evaluating the physiological quality and lignin content in the seed coat of the conventional and RR soybean seeds collected at three different times in Lavras (MG), Brazil. Thus, the seeds of ten cultivars collected at stages R7, R8 and 20 days of harvest delay (R8+20) were submitted to tests for evaluation of physiological quality and lignin content. Harvest stages were determined according to Fehr & Caviness (1977).

We observed differences in the physiological quality of seeds among the different harvest times for the cultivars BRS 134, BRS 247 RR, Conquista, Jataí and Silvânia RR, with reduction in viability with harvest delay (R8 + 20). In a similar way, when submitted to accelerated aging, the seeds of the cultivars BRS 245 RR, BRS 134, BRS Jataí and Silvânia RR also underwent a reduction in vigor with harvest delay (Table 1). Braccini et al. (2003), studying the response of 15 genotypes of soybeans to harvest delay, also observed a significant reduction in germination percentage and vigor of seeds when they were submitted to harvest 30 days after the R8 stage of development.


Means followed by the same letter in the line for each determination do not differ among themselves by the Scott-Knott test at the 5% significance level.

Table 1. Means of the germination and accelerated aging test (% of normal seedlings) from seeds of soybean cultivars and their respective RR genetically modified forms, 2007/08 harvest. UFLA, Lavras, MG, Brazil.

We observed that the greatest decreases in seed vigor by the accelerated aging test (Table 1), when the harvest delay and the mean of stages R7 and R8 are contrasted, occurred for the

impermeability to water as an alternative for avoiding loss of quality in the field (Gilioli & França Neto, 1982; Peske & Pereira, 1983; Hartwig & Potts, 1987), with delay in harvest and determination of the lignin content in the seed coat being methodologies suggested for genetic breeding programs for evaluation of the quality of soybean seeds (França Neto &

Within this context, the work presented below (Gris et al., 2010) was conducted with the purpose of evaluating the physiological quality and lignin content in the seed coat of the conventional and RR soybean seeds collected at three different times in Lavras (MG), Brazil. Thus, the seeds of ten cultivars collected at stages R7, R8 and 20 days of harvest delay (R8+20) were submitted to tests for evaluation of physiological quality and lignin content.

We observed differences in the physiological quality of seeds among the different harvest times for the cultivars BRS 134, BRS 247 RR, Conquista, Jataí and Silvânia RR, with reduction in viability with harvest delay (R8 + 20). In a similar way, when submitted to accelerated aging, the seeds of the cultivars BRS 245 RR, BRS 134, BRS Jataí and Silvânia RR also underwent a reduction in vigor with harvest delay (Table 1). Braccini et al. (2003), studying the response of 15 genotypes of soybeans to harvest delay, also observed a significant reduction in germination percentage and vigor of seeds when they were

Celeste 94.75a 96.50a 95.50a 94.75a 97.50a 91.50a 77.01a 82.42a 94.76a Baliza RR 94.25a 93.00a 91.00a 91.50a 88.50a 84.00a 83.47b 90.86b 118.01a BRS 133 91.25a 93.00a 88.00a 91.25a 96.50a 87.50a 93.66b 82.61b 107.15a

RR 91.75a 96.50a 90.50a 97.75a 99.50a 87.50b 94.79a 99.11a 97.37a BRS 134 91.75a 90.50a 79.00b 94.00a 95.00a 75.50b 86.81a 87.02a 97.65a

RR 96.50a 98.50a 87.50b 94.00a 96.00a 87.00a 76.38a 85.18b 102.44b Conquista 85.75a 90.00a 78.00b 89.75a 88.50a 84.00a 93.87b 85.23b 118.25a Valiosa RR 89.75a 83.00a 84.50a 87.50a 92.00a 87.50a 98.15b 90.01b 112.56a Jataí 91.50a 89.00a 76.50b 93.25a 87.00a 64.00b 83.42b 88.43b 152.70a

RR 93.00a 91.50a 82.00b 92.50a 92.00a 71.00b 92.92b 89.61b 143.74a Means followed by the same letter in the line for each determination do not differ among themselves by

Table 1. Means of the germination and accelerated aging test (% of normal seedlings) from seeds of soybean cultivars and their respective RR genetically modified forms, 2007/08

We observed that the greatest decreases in seed vigor by the accelerated aging test (Table 1), when the harvest delay and the mean of stages R7 and R8 are contrasted, occurred for the

Germination Accelerated Aging Electrical Conductivity R7 R8 R8 + 20 R7 R8 R8 + 20 R7 R8 R8 + 20

Harvest stages were determined according to Fehr & Caviness (1977).

submitted to harvest 30 days after the R8 stage of development.

Krzyzanowski, 2003).

Cultivars

BRS 245

BRS 247

Silvânia

the Scott-Knott test at the 5% significance level.

harvest. UFLA, Lavras, MG, Brazil.

cultivars Jataí and Silvânia RR, which presented, on average, losses in vigor of 40.82% and 29.93% respectively, indicating that not always cultivars that have high seed quality when collected near physiological maturity have greater tolerance to deterioration with delay of harvest. And, moreover, the greatest values of electrical conductivity were observed for the majority of seeds of the cultivars collected 20 days after the R8 stage, with exception of the cultivar BRS 247 RR, in which reduction in seed vigor was observed as of the R8 stage, and of the cultivars Celeste, BRS 245 RR and BRS 134 that did not undergo any alterations with the time of harvest.

As degradation of the cellular membranes is constituted hypothetically in the first event of the deterioration process (Delouche & Baskin, 1973), tests that evaluate membrane integrity, such as the electrical conductivity test, would theoretically be the most sensitive for estimating seed vigor, which is in agreement with the results obtained in this study, in which said test stood out in detecting differences of viability between the harvest times in seven of the ten cultivars evaluated. We emphasize that the electrical conductivity values observed in this study were situated from 77.01 μS cm-1 g-1 to 98.15 μS cm-1 g-1 for the R7 harvest time, 82.42 μS.cm-1.g-1 to 99.11 μS.cm-1.g-1 for the R8 harvest time and 94.76 μS.cm-1.g-1 and 152.70 μS.cm-1.g-1 for the 20 days after R8, values which demonstrate the growing trend of leachates released by the seeds with delay in harvest.

When we analyze the percentage of mechanical damage in seeds (Table 2), we observe the greatest values with delay of harvest for the cultivars Conquista (12.5%), Jataí (16.0%) and Silvânia RR (15.0%), which was not observed for the other cultivars studied. In addition, we also observed that by the germination test of seeds submitted to the water immersion test, three of the ten cultivars evaluated were differentiated in regard to the percentage of normal seedlings, however, with distinct responses. The lowest germination values when collected in R8 were observed in seeds of the cultivar BRS 245 RR; in those of the cultivar BRS 247 RR there was a reduction in germination when collected in R8 and R8 + 20; and finally in those of the cultivar Silvânia RR the lowest germinative power was verified when collected in R7 and R8. Various authors emphasize that soybean cultivars and lines behave differently in regard to degree of tolerance to delay of harvest (Lin & Severo, 1982; Rocha, 1982; Boldt, 1984), indicating that this trait may influence maintenance of the physiological quality of the seeds.

For the lignin content in the soybean seed coat, we can observe greater lignin content in the seed coat of seeds collected in the R7 and R8 + 20 stages, as well as for the cultivar Silvânia RR, when contrasted with its conventional version Jataí (Table 3).

When we observe the data of percentage of deformed abnormal seedlings, characterized by root curling, typical of damage by rapid imbibition, we observe a smaller number of abnormal seedlings due to the greater number of dead seeds with harvest delay. Giurizatto et al. (2003) affirm that the deteriorated seeds imbibe more rapidly and are therefore more prone to greater damage through imbibition, which is in agreement with the results obtained in this study.

According to Alpert & Oliver (2002) the cellular membranes have two main states, one more fluid or "crystalline liquid" and another less fluid or "gel", remaining, when organized, in the crystalline phase. In a dry seed, the membranes are found in the gel phase and therefore do not constitute an efficient barrier to contain the release of solutes. When the seeds are exposed to rapid imbibitions, the water penetrates before the membrane can be reverted to the crystalline liquid phase, with damage occurring to the cells; thus, the transition between these two phases in the configuration of the membrane constitutes the fundamental cause of possible injuries during imbibition of seeds, which makes the study of the role of lignin in the seed coat even more important.


Means followed by the same letter in the line for each determination do not differ among themselves by the Scott-Knott test at the 5% significance level.

Table 2. Means obtained for mechanical damage (%) and germination after water immersion (% of normal seedlings and abnormal curled) of soybean cultivar seeds and their genetically modified RR forms, 2007/08 harvest. UFLA, Lavras, MG, Brazil.

These differences observed for the lignin content among the harvest times are not biologically explainable, having possibly been detected due to the low coefficient of variation (CV) obtained for this variable. When we analyze the sole significant contrast, for its part, the genetically modified cultivar Silvânia RR presented greater lignin content in the seed coat than its respective conventional cultivar Jataí. Nevertheless, as an isolated fact, among the five RR combinations versus the conventional versions tested, in our view it does not justify a greater inference regarding pleiotropy of the RR transgene.


Means followed by the same letter in the column do not differ among themselves by the Scott-Knott test at the 5% significance level.

Table 3. Means with a significant difference obtained for lignin content in the soybean seed coat (%), 2007/08 harvest, UFLA, Lavras, MG, Brazil.

In general, we can conclude that in spite of there being behavioral differences in regard to tolerance to harvest delay among the different cultivars evaluated, we did not observe consistent results in regard to a comparison of the RR versus conventional cultivar, not indicating, for the conditions of this test, any sign of pleiotropy.

Celeste 3.50a 2.50a 3.00a 62.50a 70.50a 62.00a 20.00a 14.50a 11.00a Baliza RR 3.00a 3.00a 6.00a 50.00a 46.50a 44.50a 17.50a 24.00a 12.00a BRS 133 3.00a 1.00a 2.00a 55.00a 49.50a 43.50a 14.50b 26.00a 17.00b BRS 245 RR 2.50a 2.50a 5.00a 46.00a 22.50b 43.50a 19.00a 17.00b 32.00b BRS 134 1.50a 1.50a 1.00a 51.00a 47.50a 36.00a 26.00a 26.50a 23.00a BRS 247 RR 1.50a 1.00a 3.50a 63.00a 50.50b 41.00b 15.50b 32.00a 23.00b Conquista 6.00b 4.50b 12.50a 38.00a 33.00a 35.00a 10.50a 6.00a 1.00a Valiosa RR 5.50a 4.50a 5.50a 35.50a 25.50a 36.00a 9.00a 4.50a 3.50a Jataí 2.50b 3.50b 16.00a 20.50a 29.50a 26.00a 32.00a 41.50a 2.50b Silvânia RR 4.50b 5.00b 15.00a 21.50b 28.50b 40.50a 30.00a 26.00a 1.50b Means followed by the same letter in the line for each determination do not differ among themselves by

Table 2. Means obtained for mechanical damage (%) and germination after water immersion (% of normal seedlings and abnormal curled) of soybean cultivar seeds and their genetically

These differences observed for the lignin content among the harvest times are not biologically explainable, having possibly been detected due to the low coefficient of variation (CV) obtained for this variable. When we analyze the sole significant contrast, for its part, the genetically modified cultivar Silvânia RR presented greater lignin content in the seed coat than its respective conventional cultivar Jataí. Nevertheless, as an isolated fact, among the five RR combinations versus the conventional versions tested, in our view it does

Lignin Content Harvest Stages Cultivars R7 R8 R8 + 20 Jataí Silvânia RR 0.2685a 0.2385b 0.2615a 0.3008b 0.4167a Means followed by the same letter in the column do not differ among themselves by the Scott-Knott test

Table 3. Means with a significant difference obtained for lignin content in the soybean seed

In general, we can conclude that in spite of there being behavioral differences in regard to tolerance to harvest delay among the different cultivars evaluated, we did not observe consistent results in regard to a comparison of the RR versus conventional cultivar, not

Germination after Immersion Normal seedlings

R7 R8 R8 + 20 R7 R8 R8 + 20 R7 R8 R8 + 20

Germination after Immersion Abnormal curled seedlings

Cultivars

Mechanical Damage

the Scott-Knott test at the 5% significance level.

at the 5% significance level.

modified RR forms, 2007/08 harvest. UFLA, Lavras, MG, Brazil.

not justify a greater inference regarding pleiotropy of the RR transgene.

coat (%), 2007/08 harvest, UFLA, Lavras, MG, Brazil.

indicating, for the conditions of this test, any sign of pleiotropy.

#### **3.2 Physiological quality and lignin content in the plants submitted to spraying with glyphosate**

Glyphosate (N-phosphonomethyl glycine) is one of the most used herbicides in weed control throughout the world, making up nearly 12% of global herbicide sales and presenting more than 150 commercial brands (Kruse et al., 2000). The emergence of RR genetically modified soybeans increased the use of this molecule in soybeans crops in a considerable way and, along with this, also the environmental concern due to exclusive and indiscriminate use of this herbicide.

According to Sanino et al. (1999), although pesticides (especially glyphosate) may have a beneficial effect on agricultural productivity, the potential risk of these chemical compounds in the environment must be considered, which makes greater studies regarding the behavior of glyphosate under tropical conditions relevant. Within this context we aimed to evaluate the physiological quality of genetically modified RR soybean seeds and the lignin contents of plants submitted to spraying with glyphosate herbicide (Gris, 2009).

In Tables 4 and 5 we present the mean results for the variables analyzed when the soybean plants were submitted to spraying with glyphosate herbicide and water (greenhouse test) and spraying with glyphosate herbicide or manual weeding (field test) respectively.


Means followed by the same letter in the line for each determination do not differ among themselves by the Scott-Knott test at the 5% significance level.

Table 4. Means of germination and Accelerated aging (% of normal seedlings) and lignin content in the seed coat (%) of genetically modified RR soybean seeds submitted to spraying with water and glyphosate herbicide, 2007/08 harvest, Lavras, MG, Brazil, greenhouse test.

We observe that application of the glyphosate herbicide did not alter the physiological quality of the soybean seeds nor the lignin contents in the seed coat and in the plant for the two tests evaluated. These results are not in agreement with those obtained by Sanino et al. (1999), who studying the effect of application of glyphosate herbicide in soybeans observed, in a general way, reduction in the physiological quality of the RR seeds, as well as considerable reduction in activity of the enzyme α-amylase in terms of time. It is worth emphasizing that such a study was carried out comparing only 2 soybean cultivars, one conventional and one genetically modified RR variety, and that the two did not represent the same genotype, since they originated from different parentages.

In this study (Gris, 2009) we obtained a significant response only for the interaction cultivar versus treatments, when the values of electrical conductivity of the seeds produced in the field test were evaluated (Table 5), in which we observed that seeds of the cultivars Baliza RR and BRS 247 RR had their values reduced and increased respectively when the same spraying was performed. Such a differential response may possibly be explained by the different capacity of the genes inserted in the RR cultivars in expressing tolerance to the glyphosate herbicide, which according to Lacerda & Matallo (2008) may or may not occur in a homogeneous manner among cultivars and even within the same cultivar, as well as other factors inherent to the genetics of each cultivar.


Means followed by the same letter in the line for each determination do not differ among themselves by the Scott-Knott test at the 5% significance level.

Table 5. Means of germination and Accelerated aging (% of normal seedlings), Mechanical damage (%), Emergence speed index – ESI (days), Electrical conductivity (µS.cm-1.g-1), Lignin content in the seed coat, pod and stem (%) of genetically modified RR soybean cultivars submitted to manual weeding and spraying with glyphosate herbicide, 2007/08 harvest, Lavras, MG, Brazil, field test.

It is worth emphasizing that since degradation of the cellular membranes is constituted hypothetically in the first event of the deterioration process (Delouche & Baskin, 1973), tests such as electrical conductivity that evaluate membrane integrity are theoretically most sensitive for estimating seed vigor, which possibly, allied with the affirmations of Lacerda & Matallo (2008), would explain the alterations only in the conductivity values.

The absence of a significant response for treatments with weeding and spraying with the glyphosate herbicide indicate that in a general way they did not influence the physiological quality of the seeds, nor the lignin content in the soybean plants. According to Cole & Cerdeira (1982) the blocking of the shikimate pathway due to the action of the glyphosate leads to the accumulation of shikimic acid with many physiological and ecological implications, which, according to Duke & Hoagland (1985) and Becerril et al. (1989), may result in synthesis of indol acetic acid of other plant hormones, chlorophyll synthesis, phytoalexin and lignin synthesis and protein synthesis, and affect photosynthesis, respiration, transpiration, permeability of membranes and other factors.

In addition, other studies have shown that applications of glyphosate in crops interfere in nutrient absorption, increase pests and diseases, reducing crop vigor and yield (Antoniou et al., 2010). According to compilation of data made by these authors, glyphosate reduces nutrient absorption by plants, immobilizing trace elements such as iron and manganese in the soil, as well as avoiding their transport from the roots to the above ground part

Baliza RR 93.50a 96.50a 88.00a 90.50a 0.75a 0.75a 7.07a 7.14a BRS 245 RR 89.00a 93.50a 97.00a 95.50a 1.00a 0.00a 7.24a 7.12a BRS 247 RR 96.50a 97.75a 94.50a 91.50a 0.75a 1.75a 7.19a 7.07a Silvânia RR 93.00a 93.00a 87.00a 88.00a 3.00a 2.00a 7.44a 7.00a Valiosa RR 86.00a 89.00a 84.50a 83.50a 2.00a 2.25a 7.42a 7.39a

Baliza RR 61.0a 48.0b 0.24a 0.23a 8.66a 8.09a 12.71a 13.61a BRS 245 RR 69.0a 70.0a 0.19a 0.19a 8.41a 7.81a 12.20a 13.07a BRS 247 RR 52.0b 70.0a 0.20a 0.20a 9.26a 8.57a 13.43a 12.91a Silvânia RR 69.0a 69.0a 0.29a 0.30a 9.61a 9.03a 20.23a 18.50a Valiosa RR 46.0a 40.0a 0.27a 0.30a 7.87a 7.70a 14.62a 13.35a Means followed by the same letter in the line for each determination do not differ among themselves by

Table 5. Means of germination and Accelerated aging (% of normal seedlings), Mechanical damage (%), Emergence speed index – ESI (days), Electrical conductivity (µS.cm-1.g-1), Lignin content in the seed coat, pod and stem (%) of genetically modified RR soybean cultivars submitted to manual weeding and spraying with glyphosate herbicide, 2007/08

It is worth emphasizing that since degradation of the cellular membranes is constituted hypothetically in the first event of the deterioration process (Delouche & Baskin, 1973), tests such as electrical conductivity that evaluate membrane integrity are theoretically most sensitive for estimating seed vigor, which possibly, allied with the affirmations of Lacerda &

The absence of a significant response for treatments with weeding and spraying with the glyphosate herbicide indicate that in a general way they did not influence the physiological quality of the seeds, nor the lignin content in the soybean plants. According to Cole & Cerdeira (1982) the blocking of the shikimate pathway due to the action of the glyphosate leads to the accumulation of shikimic acid with many physiological and ecological implications, which, according to Duke & Hoagland (1985) and Becerril et al. (1989), may result in synthesis of indol acetic acid of other plant hormones, chlorophyll synthesis, phytoalexin and lignin synthesis and protein synthesis, and affect photosynthesis,

In addition, other studies have shown that applications of glyphosate in crops interfere in nutrient absorption, increase pests and diseases, reducing crop vigor and yield (Antoniou et al., 2010). According to compilation of data made by these authors, glyphosate reduces nutrient absorption by plants, immobilizing trace elements such as iron and manganese in the soil, as well as avoiding their transport from the roots to the above ground part

Matallo (2008), would explain the alterations only in the conductivity values.

respiration, transpiration, permeability of membranes and other factors.

Germination Accelerated Aging Mechanical Damage ESI

Electrical Conduct. Seed Coat Lignin Pod Lignin Stem Lignin

Herbicide Weeding Herbicide Weedin

Herbicide Weeding Herbicide Weedin Herbicide

g

Herbicide

Cultivars

Cultivars

Weeding Herbicide Weedin

Weeding Herbicide Weedin

the Scott-Knott test at the 5% significance level.

harvest, Lavras, MG, Brazil, field test.

g

g

(Strautman, 2007). As a result, RR soybean plants treated with glyphosate have lower levels of manganese and other nutrients and reduction in growth of budding and roots (Zobiole et al., 2010). It is worth emphasizing that the seeds produced in the two tests described in this secondary heading are being tested in regard to variation in chemical composition, data which should soon be published.

Both in the field test and in the greenhouse test, it was not possible to relate physiological quality of the seeds and lignin content in their seed coat. We observed significant differences only among the cultivars evaluated, which presented different responses when submitted to the different vigor tests, as well as lignin content, which was already expected, in terms of the great genetic variability among them.

We conclude from these tests that there is a differential response for the electrical conductivity values of the seeds when the plants of different soybean cultivars are submitted to spraying with the glyphosate herbicide; nevertheless, we did not observe a difference in the lignin contents in the stem, in the pod and in the seed coat of the soybean seeds in the cultivars evaluated when submitted to spraying with the glyphosate herbicide.

#### **3.3 Agronomic characteristics and quality of soybean seeds produced at different times**

It is known that different planting times, influenced by different environmental conditions, may be determining factors for the development of seed deterioration tolerance mechanisms and therefore for the quality of soybean seeds. Considered as a seed deterioration tolerance mechanism, the impermeability of the seed coat, characterized principally by seeds with greater lignin content, hinders water penetration in the seed coat. In a similar way to alterations in the germination process and in manifestation of vigor, in terms of the climate in the seed production phase, environmental conditions may also in some way affect the metabolism and chemical constitution of the seeds.

As we have already seen in this chapter, according to some authors, overproduction of lignin in RR soybean plants may be associated with the presence of water deficit and high temperatures during cropping, indicating that the environmental conditions found in the field during crop development may affect lignin production in the plant in an expressive way.

With this objective, we compared agronomic traits of the plant, physiological quality and seed health and lignin content in the seed coat of RR and conventional seeds produced in different time periods, summer and winter (Gris, 2009), with the determinations: plant height, height of insertion of the first pod and number of pods per plant, weight of 1000 seeds (Brasil, 1992), lignin content in the seed coat (Capeleti et al., 2005), incidence of mechanical damage (Marcos Filho et al., 1987), germination and dry matter of normal seedling from germination (Brasil, 1992), emergence speed index and germination speed index (Edmond & Drapala, 1958), final stand in the seed bed (counting at 24 days after seeding), accelerated aging at 42oC for 72h (Marcos Filho, 1999), electrical conductivity (Vieira, 1994), water immersion test of seeds and seed health, evaluating the infestation percentage (Machado, 2000) and intensity of the inoculums. The data of inoculum density were weighted by the McKinney formula (1923):

$$\Pi(\%) = \frac{\sum \left( \mathbf{F} \times \mathbf{n} \right) \times 100}{\left( N \times M \right)}$$

In which: II = inoculum intensity, F = number of seeds with a determined score, n = score observed, N = total number of seeds evaluated and M = maximum score of the scale.

In Table 6, we present a summary of the mean results for the variables in which the contrasts (RR cultivar versus conventional cultivar) presented a significant difference, for both harvests, in which among all the characteristics evaluated, significant results for the contrasts evaluated were few.

For the electrical conductivity test, we observed a greater value for the conventional cultivar Jataí (76.54 µS.cm-1.g-1) when compared to the cultivar Silvânia RR (100.25 µS.cm-1.g-1). According to Vieira & Krzyzanowski (1999) for lots of high vigor soybean seeds, the standard conductivity values should be situated at most up to 70-80 µS.cm-1.g-1, however with a strong trend to present medium vigor. Nevertheless, in spite of the high value of electrical conductivity observed in seeds of the cultivar Silvânia RR, we did not observe differences between the two cultivars in the germination and vigor tests, which, according to José et al. (2004), may indicate that there are cultivars with greater efficiency in membrane reorganization, not resulting in damages, strictly speaking.


Capital letters followed by the same letter in the line do not differ among themselves by the Scheffe Test, at the 5% significance level.

Table 6. Mean values for some variables in which the contrasts between the conventional soybean cultivar and its genetically modified RR version presented significance, summer and winter harvest, Lavras, MG, Brazil.

Panobianco (1997) upon reporting variation in electrical conductivity of soybean seeds and the lignin content in their seed coat affirms that the genotype may alter the electrical conductivity for seeds with the same standard of physiological quality. Nevertheless, we did not observe significant differences between the cultivars Jataí and Silvânia RR in regard to lignin content in the seed coat, indicating that, in this case, it may not have been responsible for the variation in electrical conductivity observed. In the same way, it was not possible to relate the difference in the lignin contents in the seed coat, observed between the cultivars Celeste (0.20%) and Baliza RR (0.26%), and the results of physiological quality of the two, produced in the summer harvest, since they differed only for this characteristic. It is worth highlighting that in spite of the differences found for these two cultivars, it was not possible through the incidence of mechanical damage to detect any differences between the cultivars studied.

Upon observing the contrasts established between the RR and conventional cultivars, we can infer that the cultivars Jataí and Silvânia RR presented the greatest number of significant differences among the variable studied (Table 6), not only in relation to the physiological quality of the seeds, but also in regard to agronomic traits, such as plant height and number of pods per plant.

In Table 6, we present a summary of the mean results for the variables in which the contrasts (RR cultivar versus conventional cultivar) presented a significant difference, for both harvests, in which among all the characteristics evaluated, significant results for the

For the electrical conductivity test, we observed a greater value for the conventional cultivar Jataí (76.54 µS.cm-1.g-1) when compared to the cultivar Silvânia RR (100.25 µS.cm-1.g-1). According to Vieira & Krzyzanowski (1999) for lots of high vigor soybean seeds, the standard conductivity values should be situated at most up to 70-80 µS.cm-1.g-1, however with a strong trend to present medium vigor. Nevertheless, in spite of the high value of electrical conductivity observed in seeds of the cultivar Silvânia RR, we did not observe differences between the two cultivars in the germination and vigor tests, which, according to José et al. (2004), may indicate that there are cultivars with greater efficiency in membrane

Variables Means – Summer 2006/07 harvest

Variables Means – Winter 2007 harvest

Capital letters followed by the same letter in the line do not differ among themselves by the Scheffe

Table 6. Mean values for some variables in which the contrasts between the conventional soybean cultivar and its genetically modified RR version presented significance, summer

Panobianco (1997) upon reporting variation in electrical conductivity of soybean seeds and the lignin content in their seed coat affirms that the genotype may alter the electrical conductivity for seeds with the same standard of physiological quality. Nevertheless, we did not observe significant differences between the cultivars Jataí and Silvânia RR in regard to lignin content in the seed coat, indicating that, in this case, it may not have been responsible for the variation in electrical conductivity observed. In the same way, it was not possible to relate the difference in the lignin contents in the seed coat, observed between the cultivars Celeste (0.20%) and Baliza RR (0.26%), and the results of physiological quality of the two, produced in the summer harvest, since they differed only for this characteristic. It is worth highlighting that in spite of the differences found for these two cultivars, it was not possible through the incidence of

Upon observing the contrasts established between the RR and conventional cultivars, we can infer that the cultivars Jataí and Silvânia RR presented the greatest number of significant differences among the variable studied (Table 6), not only in relation to the physiological quality of the seeds, but also in regard to agronomic traits, such as plant height and number

mechanical damage to detect any differences between the cultivars studied.

Plant height (m) Jataí 1.56 a vs Silvânia RR 1.41 b Number Pods/plant Jataí 110.00 a vs Silvânia RR 57.50 b Germination (%) BRS 133 95.50 a vs BRS 245 RR 87.25 b Weight of 1000 seeds (g) BRS 134 155.50 a vs BRS 247 RR 142.70 b Emergence Speed Index BRS 134 7.16 b vs BRS 247 RR 7.55 a Lignin Seed Coat (%) Celeste 0.20 b vs Baliza RR 0.26 a

(µS.cm-1.g-1) Jataí 76.54 b vs Silvânia RR 100.25 a

contrasts evaluated were few.

Electrical conductivity

Test, at the 5% significance level.

of pods per plant.

and winter harvest, Lavras, MG, Brazil.

reorganization, not resulting in damages, strictly speaking.

When we analyze the mean values of plant height and number of pods per plant, we verify once more that the conventional cultivar Jataí showed superiority to the cultivar Silvânia RR, such that for number of pods/plant, these values were up to 91.3% greater. Nevertheless, it is worth emphasizing that for these two cultivars in field conditions, we observed the greatest variations in regard to the phenological cycle, with greater uniformity in maturation and a shorter cycle, around 10 days, of the conventional cultivar Jataí in relation to the genetically modified RR cultivar. It is fitting to highlight that in spite of the RR cultivars tested in this study being essentially derivatives of the respective conventional cultivars, by means of backcrossings, the genotype of the recurrent genitor is not always recovered, due to number fewer recurrence cycles which may consequently result in variations between both materials. Nevertheless, for these cultivars, there is no information on the number of backcrossing cycles used.

When we evaluate the physiological quality of the seeds by means of the germination test in the summer harvest and of the germination speed index (IGV) in the winter harvest, we do not observe a relationship between the significant results for these variables, with the contrasts BRS 133 versus BRS 245 RR and Conquista versus Valiosa RR being differentiated respectively. For both results, the conventional cultivars showed superiority to the genetically modified RR cultivars, with the conventional cultivar BRS 133, with 95% of normal seedlings, overcoming the cultivar BRS 245 RR, with 87%, by approximately 9.5%, when they were produced in the summer harvest. Nevertheless, by the results in reference to the Emergence Speed Index, we observe a lower value for the genetically modified cultivar BRS 247 RR (7.55 days) in comparison with the conventional cultivar BRS 134 (7.16), which once more shows the inconsistency of data that justify a pleiotropic effect of the RR gene on lignin production.

It is worth emphasizing that in spite of the results found in this study, with exception of the variables Emergence Speed Index and lignin in the seed coat, the RR cultivars stood out in relation to the conventional cultivars; most of the significant contrasts, were seen to be isolated, in only one of the harvests or one of the tests in the midst of various comparisons among physiological quality of the seeds, therefore not indicating substantial differences of quality between the RR and conventional materials.

According to Menezes (2008) the physiological quality of soybean seeds is influenced by the maternal or extra-chromosome effect, just as is the cytoplasmatic inheritance, with the physical characteristics of the seed coat, of maternal origin, not being sole determinants of the physiological quality of the seeds. According to this author, the study of genetic control for seed quality indicates the effect of the general and specific combination capacity, which suggests the presence of additive and non-additive gene effects for physiological quality of soybean seeds. Therefore, the quality of seeds may not be attributed only to their seed coat and consequently to their lignin contents, but also to genes present in the nucleus.

When we analyze the results obtained in the seed health test (Figure 6), we observe that the cultivars BRS 133, BRS 245 RR, BRS 134 and BRS 247 RR presented the lowest percentages of infection and infection indexes (severity), when produced in the summer, indicating that the environmental conditions during the seed maturation period were responsible for seed health quality. In these cultivars a shorter phenological cycle and semi-early maturity was observed, which provided for the maturation period outside of the rainy period.

According to Delouche (1975), the alternating of dry and wet days during the maturation phase until harvest, which occurs with greater facility in the summer, can increase the incidence of diseases in a differentiated way at the end of the cycle of the seeds produced. Within this context, the seed becomes not only an easy target for the action of microorganisms, which considerably reduce its viability, but they also come to be efficient vehicles for dissemination of pathogens (Machado, 2000). This situation may be visualized principally for the cultivars Jataí and Silvânia RR, which remained for a greater period in the field, and presented the greatest percentages of infection, 39% and 38% (Figure 6A), and also the greatest indexes of infection by the pathogen Phomopsis, 35% and 26% (Figure 6B), respectively. It is worth emphasizing that when produced in winter conditions, under a controlled irrigation system, without rains in the seed maturation period, the presence of pathogens was not observed for any seeds.

Fig. 6. Average values for infection percentage (A) and infection indexes - severity (B) in the seed health test of conventional soybeans and the genetically modified RR versions, summer harvest.

In relation to the RR versus conventional contrasts, we observe that in spite of the cultivars tested in this study having their origin in the same genotype by successive backcrossings, when observed in the field, we verified that some presented perceptible cycle variations, maintaining the cultivars Conquista and Celeste for more days in the field in relation to the cultivars Valiosa RR and Baliza RR, respectively; enough so that the first, subjected to rains and high temperatures, presented slightly greater values in the seed health and severity test. In this case, we cannot attribute the differences of RR versus conventional contrast, observed in Figure 6A and 6B, to the effect of the RR transgene, but rather to environmental conditions associated with difference of cycle.

In view of the above, in spite of some authors suggesting the pleiotropic effect of the transgene CP4 EPSPS on lignin overproduction in the plant, it was not possible for us to identify the pleiotropic effect in the cultivars studied in this and in the other studies described here, which indicates that the alterations of lignin content in the plant, observed by those authors under normal climatic conditions, are not due to the fact of the lignin molecule precursors being formed in the shikimic acid pathway. Thus, the sequence CP4 EPSPS, introduced in the genome of commercial soybean cultivars, responsible for the production of the protein CP4 enolpyruvylshikimate-3-phosphate-synthase (EPSPS), an enzyme that participates in the biosynthesis of aromatic amino acids in plants and microorganisms, seems not to be associated with lignin contents in the plant and in the soybean seed coat, and it seems that there are no substantial differences in regard to the agronomic traits and physiological quality of seeds between conventional and genetically modified RR cultivars.
