| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Institute of Crop and Grassland Science, Federal Agricultural Research Center, 38116 Braunschweig, Germany
b Institute of Integrated Plant Protection, Federal Biological Research Center for Agriculture and Forestry, 14532 Kleinmachnow, Germany
c Institute for Plant Virology, Microbiology and Biosafety, Federal Biological Research Center for Agriculture and Forestry, 38104 Braunschweig, Germany
d Institute of Agricultural Crops, Federal Center for Breeding Research on Cultivated Plants, 18190 Groß Lüsewitz, Germany
* Corresponding author (gerhard.ruehl{at}fal.de).
 |
ABSTRACT |
|---|
 TOPNOTES ABSTRACT INTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Abbreviations: BS, Braunschweig, CG, clover–grass • DD, Dahnsdorf • EC, European Communities • GM, genetically modified • MS, Mariensee • nt, nucleotide • PCR, polymerase chain reaction • rt PCR, real-time PCR • SF, sunflower • WH, Wendhausen
|
ACKNOWLEDGMENTS |
|---|
|
NOTES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Received for publication July 18, 2007.
a Institute of Crop and Grassland Science, Federal Agricultural Research Center, 38116 Braunschweig, Germany
b Institute of Integrated Plant Protection, Federal Biological Research Center for Agriculture and Forestry, 14532 Kleinmachnow, Germany
c Institute for Plant Virology, Microbiology and Biosafety, Federal Biological Research Center for Agriculture and Forestry, 38104 Braunschweig, Germany
d Institute of Agricultural Crops, Federal Center for Breeding Research on Cultivated Plants, 18190 Groß Lüsewitz, Germany
* Corresponding author (gerhard.ruehl{at}fal.de).
One approach to ensuring coexistence of genetically modified (GM) and conventional maize (Zea mays L.) is reducing pollen-mediated gene flow. Field experiments were conducted in 2005 at four sites in Germany to compare a tall sunflower crop (Helianthus annuus L.) vs. a short clover–grass crop (Trifolium pratense L. and Lolium spp.) with regard to their ability to reduce outcrossing when grown as buffer between pollen donor and recipient maize plots. Three different maize test systems were used: (i) quantification of a donor transgene via real-time polymerase chain reaction (rt PCR), (ii) a nontransgenic test system based on a dominant kernel color trait, and (iii) a molecular marker test system based on rt PCR quantification of a cultivar-specific nontransgenic DNA sequence. We found that the three test systems yielded comparable results concerning buffer-crop effectiveness and edge effects. There was no difference in outcrossing rates when comparing the sunflower vs. clover–grass buffer crop. Outcrossing rates downwind beyond 12 m sunflower as buffer crop within adjacent 12-m-wide recipient maize were 4.2, 11.7, and 3.8% for the GM maize, the kernel color, and the molecular marker test system compared with clover–grass with 4.3, 9.6, and 3.6%. Pronounced edge effects were detected at the edges of recipient maize fields. Based on the present study, growing sunflower as a tall crop between GM and non-GM maize cannot be recommended as an appropriate coexistence measure.
Abbreviations: BS, Braunschweig, CG, clover–grass • DD, Dahnsdorf • EC, European Communities • GM, genetically modified • MS, Mariensee • nt, nucleotide • PCR, polymerase chain reaction • rt PCR, real-time PCR • SF, sunflower • WH, Wendhausen
|
INTRODUCTION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
The Commission of the European Communities (EC) fixed a mandatory labeling threshold of 0.9% for technically unavoidable ("adventitious") presence of GM material in non-GM food and feed (European Communities, 2003b,c). To keep the GM content in non-GM food and feed below the labeling threshold, the EC law requires measures to ensure the coexistence of GM crops with conventional and organic crops, which have to be implemented by the member states (European Communities, 2003a,b). In 2003, the EC proposed a catalog of measures to facilitate coexistence at the farm level (European Communities, 2003a). The most discussed items in view of central European growing conditions are isolation distances and pollen barriers (Devos et al., 2005). As in the case of certified seed production, separation of GM and non-GM maize by distance is widely regarded as a suitable tool to allow coexistence on a regional scale (Ingram, 2000; Halsey et al., 2005). In seed production, border rows of the male parent planted around the seed field act as a physical barrier to foreign pollen influx and enhance biological isolation due to competing pollen supply (Ireland et al., 2006). The efficacy of border rows was corroborated by studies on gene flow from transgenic to conventional varieties of the same species that showed increased levels of outcrossing if both crops were separated by bare ground than by the same area planted with the same species as the crop (Jones & Brooks, 1952; Morris et al., 1994; Reboud, 2003). Physical isolation can also be achieved by growing vegetative barriers that differ from the field crop. In beet seed production, for example, hemp strips are used as pollen barriers; because of their dense growth and height of up to 4 m, they prevent contamination by incoming pollen (Saeglitz et al., 2000). Jones and Brooks (1952) demonstrated the effectiveness of a row of trees and underbrush in reducing gene flow in maize. However, so far, little information from field studies exists concerning the efficacy of different vegetative barriers as coexistence measure for the reduction of outcrossing in a wind-pollinated crop species like maize.
The current study on the coexistence of maize in Germany was initiated in 2005. It aims generally to broaden the knowledge on gene flow as well as to set up a reliable database for adopting coexistence measures applicable to maize cultivation in central Europe.
Two specific objectives underlie the present study. The first was to assess the effect of clover–grass (CG) (Trifolium pratense L. and Lolium spp.) as a short buffer crop in comparison to sunflowers (SF) (Helianthus annuus L.) as a tall buffer (barrier) crop on outcrossing rates in maize to determine whether the use of buffer crop species lowers isolation distances for coexistence of maize. The second objective was to analyze the extent of gene flow into donor-facing field edges of maize recipient plots. Three different test systems were used. One of these was based on transgenic maize, and two systems used real-time (rt) polymerase chain reaction (PCR) for the quantitation of gene flow, whereas the third made use of a phenotypic marker.
|
MATERIAL AND METHODS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
To synchronize flowering of donor and recipient maize in each test system, cultivars of a similar ripening group were used and sown on the same day.
MON810 Test System
The MON810 test system used the transgenic hybrid PR39V17 as pollen donor and the isogenic non-GM cultivar Sandrina as recipient (both cultivars kindly provided by Pioneer, Buxtehude, Germany). Plants of PR39V17 carry the MON810 transgenic event and express the Bacillus thuringiensis Berliner (Bt) cry1Ab toxin, which is highly insecticidal to the European corn borer, Ostrinia nubilalis (Hübner), an important lepidopteran pest of maize in southern and central Europe. The percentage of MON810-specific DNA relative to an endogenous maize gene in each sample was quantified by rt PCR. Real-time PCR was done by a DIN EN ISO/IEC 17025 accredited laboratory (JenaGen GmbH, Jena, Germany) according to a validated MON810 event-specific protocol by Kuribara et al. (2002) and Shindo et al. (2002). The sequences of primer pairs and probes for the MON810 target sequence [intron 1 from maize hsp70 gene and synthetic cryIA(b) gene] and for the endogenous reference gene (maize starch synthase IIb gene) are given in Kuribara et al. (2002). The rt PCR was performed on an ABI Prism 7000 Sequence Detection System (Applied Biosystems, Darmstadt, Germany).
Kernel-Color Test System
The yellow-kernelled hybrid cultivar NK Delitop (Syngenta Seeds GmbH, Bad Salzuflen, Germany) was chosen as a donor, and the white-kernelled cultivar DSP17007 (Delley Samen und Pflanzen AG, Delley, Switzerland) was used as recipient. Yellow-kernel color is dominant over white-kernel color. Numbers of yellow kernels developing on white-kernel maize ears were counted, and the percentage of yellow kernels in relation to the mean total kernel number of a white-kernel maize ear was calculated. For this purpose, the average number of kernels per white-kernel maize ear was estimated for each field trial site from random samples of 150 (Braunschweig site) and 51 (Dahnsdorf site) ears, respectively.
Molecular-Marker Test System
The molecular-marker test system used the non-GM cultivar NK Bull as donor and cultivar NK Cool as recipient (both kindly provided by Syngenta Seeds GmbH, Bad Salzuflen, Germany). NK Bull differs from NK Cool by single nucleotide polymorphisms at nucleotide (nt) 194 in exon 1 as well as nt 275 and 276 in the first intron of the caffeoyl-CoA 3-O-methyltransferase 2 gene (CCoAOMT2; Guillet-Claude et al., 2004; Hackauf and Wehling, Groß Lüsewitz, personal communication, 2005). A DIN EN ISO/IEC 17025 accredited laboratory (Planton GmbH, Kiel, Germany) was assigned to develop and validate a donor-allele specific rt PCR procedure based on CCoAOMT2. The sequences of primer pairs and probes for the detection of the molecular marker were GP1bleGV251 forward primer (5'-CCA CTC CGA GGT CGG G), GP2bleGV252 backward primer (5'-CCG CGG TCA GAG ACT AAG), and GS1bleGV264 probe (FAM-TAC CAG GTA AAC AAG CTG AGC GCA ATG AG-TAMRA). Maize alcohol dehydrogenase was used as endogenous reference gene; the primer pairs and probes were GP1GV117 forward primer (CCA GCC TCA TGG CCA AAG), GP2GV118 backward primer (CCT TCT TGG CGG CTT ATC TG), and GSGV119 probe (TTA GGG GCA GAC TCC CGT GTT CCC). The PCR was performed on an ABI Prism 7700 (Taqman) Sequence Detection System (Applied Biosystems, Darmstadt, Germany).
Experimental Field Trials
Gene flow between donor and recipient maize plots was assessed in 2005 at four sites in the northern part of Germany at experimental stations of the Federal Agricultural Research Center in Braunschweig (BS) and Mariensee (MS) and the Federal Biological Research Center for Agriculture and Forestry in Dahnsdorf (DD) and Wendhausen (WH) (Table 1
). Distances between the four experimental sites ranged between 15 and 225 km. Originally, a field trial at a fifth location at Dedelow in the northeast of Germany was included in the study but had to be discarded later because of extreme drought that severely compromised development of the sunflower and maize plots. All experimental fields are located in plain areas, with a maximum 15-m difference in elevation between the highest and the lowest point. At all sites, maize was grown at plant densities of 80,000 to 100,000 plants ha–1 (Table 1) and a row spacing of 0.75 m. Maize seed was protected against soil-borne pathogens by standard seed treatment fungicides (TMTD 98% Satec, a.i. thiram, Satec, Elmshorn, Germany; Maxim XL, a.i. fludioxonil and metalaxyl, Syngenta, Maintal, Germany) and by bird repellent (Mesurol, a.i. methiocarb, Bayer, Monheim, Germany). Fertilizer and postemergence herbicides were applied according to regional recommendations and taking soil tests into account. No insecticides were applied. Supplemental irrigation of the field plots was only possible in BS. Irrigation frequency and amount of irrigation water were adjusted to climatic conditions.
|
|
Weather Recording
During the maize flowering period, meteorological data were recorded by on-site weather stations. Temperature and relative humidity were recorded hourly using a thermo-hygrograph. Precipitation was measured with a Hellmann-type rain gauge. A Woelfle-type wind recorder recorded wind speed and direction at a height of 2 m (DD, MS, WH) and 10 m (BS), respectively (all meteorological instruments, Lambrecht, Göttingen, Germany). Wind data that were recorded at 10-m height were converted into 2-m values by a conversion factor of 0.773 (Loepmeier, personal communication, 2005). Meteorological data recording was done by the German National Meteorological Service.
Monitoring of Flowering
Flowering stages of both pollen donor and recipient maize hybrids were recorded over the entire flowering period between 19 July and 23 Aug. 2005 by observing 12 individual record points within the donor plots and 128, 112, 128, 96, and 130 points within the recipient plots in MS and WH (MON810), BS and DD (kernel color), and BS (molecular marker), respectively. Each record point represented 20 plants. Record points were distributed over the experimental fields in a grid the density of which was highest at the borders facing the central donor plot. Numbers of plants in full anthesis and full silking were recorded daily. At the end of the flowering period, recording was made every second day. Monitoring of flowering stages was always done between 8:00 a.m. and 1:00 p.m. Male flowers were considered to be in full anthesis if 75% of tassels were flowering (Westgate et al., 2003). The criterion for the full silking stage was complete silk exsertion. The objective of these observations was to assess coincidence of pollen shedding in the donor and silking in the recipient, as well as coincidence of anthesis in the recipient with silking in the recipient.
Sampling Strategy
To estimate gene flow in the MON810 and the molecular-marker systems, 20 ears each were collected at 533, 512, and 479 sampling points within the pollen recipient plots of sites at MS, WH (MON810), and BS (molecular marker), respectively. At these sites, ears were harvested from rows 1, 3, 6, 9, and 16 (row 14 at WH) of the inner recipient block and from rows 1, 3, 6, and 9 of the outer block. Additionally, in the outer block, rows 28 and 32 were sampled at MS, rows 16 and 30 at WH, and rows 16 and 24 at BS. This sampling scheme resulted in distances of sampling points from the pollen donor of 12 to 24 m and 36 to 60 m in case of the inner and outer block, respectively (Fig. 1). Ears were collected directly into labeled bags. After being dried down to approximately 10% moisture content in a seed dryer, each sample of 20 ears was shelled to a pooled kernel sample of approximately 6000 to 8000 seeds. One-half of each pooled sample was used for PCR analysis; the remainder acted as retain sample. Samples were stored in a seed storage room (18°C, 30% relative humidity) before PCR analysis.
To estimate the gene flow from the yellow-kernel to the white-kernel hybrid, 10 ears each were collected from 512 (BS) and 336 (DD) sample positions within the pollen recipient plots. Ears were harvested from the inner recipient block from rows 1, 3, 6, 9, 16 (as well as rows 28 and 32 in BS) and from the outer block from rows 1, 3, 6, and 9 as well as from row 32 and 40 (only BS) and from row 12 (only DD), that is, at 12 to 37 m (inner block) and at 36 to 66 m (outer block) distance from the pollen donor (Fig. 1). Visual analysis of outcrossing rates was done on the basis of undried ears within 1 to 2 d after harvest; seed color differences were stable within this period.
Sample Analysis
To optimize costs, the analyses of maize samples in the MON810 and molecular marker test system were made in two steps. First, a small number of samples distributed evenly within the recipient strips was analyzed to obtain information about the outcrossing rates within the recipient strips. On the basis of these preliminary results, a larger second set of samples corresponding to areas with the highest gene flow was selected. At least 3000 kernels were taken at random from each pooled sample to allow for the detection limit of 0.1% donor DNA at a 95% confidence interval (ISTA, 2006).
Data Analysis
Rayleigh's test was used to determine whether there was a predominant wind direction during the flowering periods at each experimental site or if wind direction was random (Batschelet, 1981).
Outcrossing data were gridded and contoured with the program Surfer version 8.00 (Golden Software Inc., Golden, CO) using the inverse distance to a power interpolation method (Davis, 1986). With this gridding method, data points are weighted by interpolation in such a manner that the influence of one point relative to another declines with distance from the grid node. For the analysis, the weighting power was set to 4.
To compare the effect of CG and SF as buffer crops on the amount of gene flow, it was necessary to identify sample points within the recipient fields with comparable donor pollen influx. Assuming a worst-case situation, downwind areas—that is, areas with maximum levels of outcrossing—were defined and just the downwind data were used for the statistical analysis. To define downwind areas in recipient fields, the mean wind direction (Table 2 ) was set as the bisecting line of a 90° angle, with the angular point sitting in the center of the pollen donor fields. All data points falling within the range of this sector were designated downwind (see Fig. 2c and d , 3c and d , 4b ) and used for the statistical analysis of the barrier-crop effect. The 90° sector was used to compensate for the variability in wind direction during flowering and to achieve equal group sizes for rows with CG and SF barrier crop. The nonparametric Wilcoxon signed rank test was applied to determine whether CG or SF caused reductions in gene flow. Mean downwind outcrossing rates per CG recipient maize row and mean downwind outcrossing rates per SF recipient maize row were compared across sites. In addition, outcrossing rates in rows with CG and SF barrier crop were compared for each experimental site separately. Comparisons were made for every row separately using the independent samples t test. In the case of unequal variances, Satterthwaite's t test was used.
|
|
|
|
Data were analyzed using SAS version 9.1 (SAS Institute, Cary, NC) and Oriana 2.0 (RockWare, Golden, CO).
|
RESULTS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
20% of donor plants were at full anthesis stage and 20% of recipient plants were at the full silking stage. Wind direction and wind speed recorded during the assumed hours of pollen shed, 6:00 a.m. to 6:00 p.m. (Henry et al., 2003; Jarosz et al., 2003; Halsey et al., 2005; own observations), are shown in Fig. 2a and b, 3a and b, and 4a. At each experimental site, there was a predominant wind direction during the flowering period (Rayleigh test, p < 0.05). The prevailing wind during flowering came from northwesterly to southwesterly directions. However, low frequencies of nonprevailing southeast wind were also recorded (Fig. 4b). During flowering, wind blew on average lightly, ranging between 1 (light air) and 2 (light breeze) on the Beaufort scale (Table 2). Mean temperature and daily precipitation that were recorded during maize flowering at each site are illustrated in Fig. 2a and b, 3a and b, and 4a. Generally, at each site, July was warmer and wetter, while August was cooler and dryer than the long-term average.
Flowering Dynamics
Maize flowering dynamics differed with the three test systems. With the MON810 test system, full silking in the recipient maize coincided with full anthesis in the recipient and full anthesis in the donor field (Fig. 2a and b). In the recipient field, the anthesis-silking synchrony lasted on average 2.9 (± 0.7 SD) days at MS and 5.3 (± 1.2 SD) days at WH. Overlap of anthesis in the donor field with silking in the recipient field was on average 3.4 (± 0.6 SD) days at MS and 5.4 (± 1.0 SD) days at WH. Flowering across individual sampling points was relatively uniform, with low deviations from the field average.
With the kernel-color test system, the anthesis-silking synchrony in the recipient field was low (Fig. 3a and b). Male and female flowers of the white-kernelled maize synchronously flowered on average 1.2 (± 0.8 SD) days at BS and 1.1 (± 1.8 SD) days at DD. At both sites, peak anthesis occurred approximately 1 wk earlier than peak silking (Fig. 3a and b). A perfect synchronization between anthesis in the donor and silking in the recipient was observed at DD (Fig. 3b). Synchronous flowering of the donor and recipient occurred on average for 5.3 (± 0.7 SD) days. At BS, peak anthesis in the donor was several days earlier than peak silking in the recipient (Fig. 3a); however, the mean anthesis-silking synchrony between donor and recipient lasted on average 3.4 (± 0.6 SD) days. However, greater deviation from the field average was observed for anthesis-silking synchrony at individual sampling points within the recipient maize fields. At BS, a higher coincidence of male and female recipient flowers was observed in the inner block compared with the outer block, whereas a more patchy distribution was found at DD.
With the molecular-marker test system, full anthesis in the recipient coincided with full anthesis in the donor field. Peak anthesis in both donor and recipient occurred 4 d earlier than peak silking in the recipient (Fig. 4a). In the recipient field, the anthesis-silking synchrony lasted on average 2.9 (± 1.0 SD) days, and the average synchrony between anthesis in the donor and silking in the recipient was 3.8 (± 1.1 SD) days. Flowering at individual sampling points did not deviate much from the field average.
General Pattern of Gene Flow
Gene flow at each site is contoured in Fig. 2c and 2d, 3c and 3d, and 4b, and mean downwind outcrossing rates per row are shown in Fig. 2e and 2f, 3e and 3f, and 4c for each test system and site. Supplemental Table 1 shows mean outcrossing rates per row that were determined north, east, south, and west of the central donor at each experimental site.
At all sites, highest levels of gene flow in the recipient maize were recorded downwind from the donor, i.e., mainly in easterly and northeasterly directions (Fig. 2c and d, 3c and d, 4b). In the downwind direction, detectable amounts of outcrossing (>0.1%) were noticed at each site up to the farthest distance investigated, that is, 54 to 66 m (Fig. 2c, 3c, 4b).
Generally, the level of downwind gene flow within both inner and outer recipient block declined rapidly with increasing distance from the donor maize (Fig. 2c and d, 3c and d, 4b).
Gene flow into recipient subplots located upwind from the pollen source was very low. In those subplots, located mainly southwesterly, outcrossing at individual sampling points in the first row of the recipient maize, that is, at 12 to 13 m from the donor, stayed below 1%. At the maximum upwind distance (54–66 m), mean outcrossing per row was 
0.1%. One exception was the kernel-color system at DD, where higher amounts of cross-fertilization were observed in the last row of the outer recipient block. At 57 m west of the pollen emitter, gene flow between individual sampling points ranged from 1.1 to 3.4%. This was most likely due to the influence of the three yellow-kernel maize fields situated westward of the experimental plot (see "Materials and Methods").
Effect of Crop Type Planted between Donor and Recipient Maize on Gene Flow
Overall, a general effect of the buffer-crop type on gene flow was not detected. Downwind outcrossing rates in specific rows within the recipient fields did not differ between CG and SF (Wilcoxon signed rank test, p always greater than 0.13).
At the single-site level, row-wise comparisons of downwind gene flow into the recipient maize across CG and SF revealed few significant differences (Fig. 2e, 3e, 4c). With the MON810 test system at the MS site, outcrossing in rows 28 and 32 of the outer block was significantly higher with the CG vs. the SF buffer crop. No significant differences, however, were found in the inner block at MS and in both the inner and outer blocks at WH (Fig. 2e and f).
With the kernel-color system at BS, significantly higher outcrossing was observed with CG vs. SF in the first recipient row facing the donor. A similar trend, although statistically insignificant, was observed at DD (Fig. 3e and f). In contrast, at the BS site, levels of downwind gene flow were significantly higher with SF in rows 16 (inner block) and 32 (outer block). No significant differences were detected at DD (Fig. 3f).
In the molecular-marker test system, significantly higher downwind outcrossing was detected with CG vs. SF barrier in rows 1 and 16 of the inner block (Fig. 4c).
Edge Effects
Across all sites, average outcrossing in the first rows of the outer blocks was higher than the average outcrossing rates in the last rows of the inner blocks, although statistically insignificant due to high variability (Wilcoxon signed rank test, p = 0.0625 for both CG and SF). Overall, a 1.8-fold (± 0.7 SD) increase in outcrossing from the last row of the inner block to the first row of the outer block was accompanied with this 12-m increase in distance from the donor when CG was planted between donor and recipient maize. With SF as a buffer crop, the average increase was even higher (2.7-fold ± 2.0 SD).
This trend was confirmed at the single-site level. Outcrossing rates in the first rows of the outer blocks were always higher in comparison to those in the last rows of the inner blocks, though in most cases not significantly. With the MON810 test system at the MS site, outcrossing rates increased from 2.9 to 3.9% (CG) and from 3.0 to 4.3% (SF) (Fig. 2e). At WH, an increase from 3.1 to 3.8% (CG) and from 2.9 to 4.0% (SF) was observed (Fig. 2f).
With the kernel-color system at BS, levels of gene flow were significantly higher in the first row of the outer block (10.41%) than in the last row of the inner block (3.64%) with CG planted between donor and recipient (Fig. 3e). A nonsignificant increase in outcrossing from the last inner row to the first outer row of 14.36 to 22.93% was observed when SF was used as buffer crop (Fig. 3e). A lower increase in outcrossing with this 12-m increase in distance from the donor was observed at the DD site. Outcrossing rates increased by 3.76% (CG) and 4.3% (SF) to give a row mean of 7.60% (CG) and 6.38% (SF) (Fig. 3f).
In the molecular-marker test system, outcrossing increased from 1.31% in the last row of the inner block to 2.02% in the first row of the outer block, with CG planted between blocks (Fig. 4c). In the case of SF, a significant increase in outcrossing from 0.16 to 0.97% was detected (Fig. 4c).
|
DISCUSSION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
One explanation for the lack of significant barrier effects of sunflower is that the mere height of a crop may not be a sufficient criterion for its ability to restrain the flow of maize pollen. Rather, plant architecture or stickiness of leafs and other plant organs may be important. Aylor et al. (2003) found that maize pollen is intercepted by the sticky trichomes of silks but not by the nonadhesive ones of maize leaves. Furthermore, they showed that pollen begins to roll off the surface of a maize leaf if the leaf-surface is inclined to at least 20° and that pollen grains can be blown off the surface of a maize leaf with wind speeds of 0.2 to 0.5 m s–1. Their results indicate that pollen dusted onto plant structures possibly can reenter the air flow. Retention of pollen by a (barrier) crop species may, thus, not only depend on its height but also on its specific surface structure.
Another possible explanation for the lack of effect on pollen retention between SF and CG is that considerable amounts of maize pollen would have been translocated by air layers moving above the tall yet relatively inflexible sunflower canopy. Regarding the short CG crop, it is assumed that the short vegetation curtailed turbulences or deflection of wind (Davis et al., 1994).
On the basis mainly of practical experience, tall plant species are used as physical barriers to pollen-mediated gene flow in seed production. For example, due to its height, a sunflower barrier has been recommended in bean (Phaseolus vulgaris L.) seed production (McCormack, 2004). In hybrid rice (Oryza sativa L.) seed production, strips of vegetative barriers 2.5 m in height and 3 to 4 m in depth have been proposed to separate the seed parents from nearby rice cultivars (Virmani et al., 2006). Moreover, physical barriers made of reed (Phragmites australis Cav.) or bast fiber are also used to reduce pollen transport by wind (Arndt and Pohl, 2005). However, scientific knowledge on the actual efficiency of these vegetative structures as barriers to pollen transport is still limited.
Several studies exist that compared the effectiveness in pollen-flow reduction of a heterospecific crop barrier as well as a zone of bare ground with a homospecific crop. One of these studies conducted in Bavaria, Germany, reported that gene flow was considerably lower when GM donor and non-GM recipient maize plots were separated by a 50-m strip of conventional maize compared with a 50-m strip of ryegrass (Lolium perenne L.) (Eder, 2006). Several authors noted that rows of maize are most effective in reducing outcrossing in maize since they function as additional pollen source that dilutes incoming foreign pollen, thereby increasing pollen competition (Jones and Brooks, 1952; Aylor et al., 2003; Goggi et al., 2006). This is supported by field studies that showed that a barren zone between donor and recipient maize is less effective in reducing gene flow compared with the same space planted with a recipient maize pollen barrier (Jones and Brooks, 1952; Wilhelm et al., 2005; Della Porta et al., 2006).
Outcrossing Rates at Field Edges (Edge Effects)
The present study demonstrates that even in the presence of a vegetative barrier downwind outcrossing rates at field edges of recipient maize fields can be higher than those measured at a smaller distance from the pollen source within a recipient field. Gene flow into the first rows of the outer blocks was higher, although in most cases not significantly, than into the last rows of the inner blocks, although the latter were 12 m closer to the pollen source. This observation was consistent across all field sites and test systems. Earlier studies described edge effects at the field margins of maize recipient plots separated from a pollen source by a nonmaize crop (Eder, 2006) or bare ground (Wilhelm et al., 2005; Della Porta et al., 2006). In the latter study, although gene flow into recipient maize fields was reduced by increasing isolation distances, outcrossing at field edges was always considerably higher compared with that at equidistant sampling points within an adjacent continuous field (Della Porta et al., 2006). These results are corroborated by small-scale field experiments by Wilhelm et al. (2005). There are two main reasons for higher outcrossing rates at field edges. First, the nonmaize buffer crop does not contribute to pollen competition between recipient and donor pollen; thus, downwind from its source, the latter can reach the silks of the recipient without being diluted by recipient pollen. Second, the field edge of a maize field may act as a windbreak, altering wind velocity and direction (Du et al., 2001) and resulting in the release of wind-transported pollen at the field edge.
The findings of the present study, in which a cropping area of conventional maize was interspersed with strips of nonmaize buffer crops, indicate that edge effects may contribute considerably to the average outcrossing rate over the whole field, especially in agriculture that is dominated by small fields and small landscape elements. However, since the amount of outcrossing decreases with increasing distance from the edge, discarding the first few rows may be an appropriate strategy to essentially reduce the GM proportion in the conventional maize harvest (Henry et al., 2003; Weber et al., 2007).
Suitability of Test Systems for Studies of Gene Flow in Maize
In the present study, outcrossing was investigated using three different test systems, one using GM and two using non-GM maize. Generally, each of the systems gave consistent and reliable results on gene flow with the specific pair of donor–recipient maize cultivars. By simultaneously planting donor and recipient maize, synchronization of anthesis in the pollen donor and silking in the recipient was possible, and thus, a prerequisite for cross-pollination was given at each site. On average, the full flowering stages of male donor and female recipient flowers overlapped for more than 3 d. However, with the kernel-color test system, we observed low coincidence of anthesis and silking within the recipient plots. As a consequence, competition between donor and recipient pollen may be low, making the white-kernelled plants of the recipient plot more receptive to foreign pollen and the test system prone to overestimation of what is expected in agricultural practice (Aylor et al., 2003; Sanvido et al., 2005).
Another consideration is the genetics of the test systems in relation to agricultural practice. The pollen donor used in the MON810 and molecular-marker systems are hemizygous and heterozygous, respectively, for the investigated trait. Consequently, the proportion of the emitted donor pollen that contributes to outcrossing in GM maize is not more than 50%. Thus, these two test systems appear well suited to reflect the gene flow that will occur in practice where hybrids hemizygous for a transgene will present the prevailing type of GM maize cultivars in the near future. In contrast, the yellow color in the kernel-color test system is a homozygous trait; that is, 100% of the pollen carries the yellow-color genes. Not surprisingly, when using this test system, measured outcrossing rates were higher than those obtained with the two other systems. A recent study showed that outcrossing rates based on the phenotypic quantification of yellow kernels in a white-kernelled hybrid were twice as high as those obtained by rt PCR–based quantification of the MON810 event (Pla et al., 2006).
Gene Flow
Maize is a wind-pollinated crop; as expected outcrossing rates at each field trial site were found to be considerably higher in areas downwind from the pollen source. Thereby, the results of the current study confirm the findings of several other field studies (Jemison and Vayda, 2001; Ma et al., 2004; Goggi et al., 2006) and demonstrate the impact of wind direction on pollen-mediated gene flow in maize. However, as discussed by Halsey et al. (2005), because of the highly complex process of gene flow, the main direction of outcrossing cannot be predicted reliably by wind measurements alone. In the current work, downwind sectors in the recipient fields were defined to identify sampling points with equal donor pollen influx representing a worst-case situation. However, in contrast to the MON810 and the molecular-marker test system, the pattern of cross-fertilization at the kernel-color sites showed that the defined downwind area did not always coincide with those parts of the recipient fields in which maximum outcrossing was actually measured (Fig. 3c and d). Most likely, gusts of wind during the silking period of the recipient maize may have additionally affected direction of pollen mediated gene flow (Ma et al., 2004; Halsey et al., 2005). Outcrossing events on the westward side of the pollen donor in BS support this assumption. Due to the hourly resolution of wind recording, the effect of gusts and nonprevailing wind on outcrossing could not be addressed in this study.
Outcrossing rates decreased with increasing distance from the pollen source within maize recipient plots. These results agree with those from previous experimental field studies (Ma et al., 2004; Goggi et al., 2006; Messeguer et al., 2006). The experimental field trials in the present study were specifically designed for the purpose of comparing the efficiency of buffer crops in reducing gene flow in maize. For this reason, the data obtained with this experimental design are not considered for discussion on minimum separation distances in maize growing. This latter aspect will be considered in a separate report based on more practice-oriented field studies.
|
CONCLUSIONS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
The study also shows that there may be pronounced edge effects, that is, high outcrossing rates at the edges of maize fields that may disproportionately contribute to the overall GM content of the harvest. As a general consequence, separate harvesting of the first GM-facing rows of conventional maize fields can be recommended as a means to decrease GM maize content of non-GM maize fields.
Probably due to homozygosity of the phenotypic marker and protandrous flowering characteristics, the phenotypic kernel-color test system used in the present study gave outcrossing rates higher than those measured with the MON810 and the molecular-marker test systems. Since modern maize hybrids generally display reduced protandry, the kernel-color test system based on the protandrous white-kernelled hybrid DSP17007 is deemed to be of limited practical relevance for studies on coexistence in maize.
The authors thank the German National Meteorological Service (DWD) for recording climate data at the experimental sites.
All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Permission for printing and for reprinting the material contained herein has been obtained by the publisher.
Received for publication July 18, 2007.
|
REFERENCES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
ková, V. Renczésová, S. Natarajan, R. Pálffy, R. Gardlík, J. Hodosy, M. Behuliak, B. Vlková, G. Minárik, T. Szemes, S. Stuchlík, and J. Tur
a. 2005. Biological and biomedical aspects of genetically modified food. Biomed. Pharmacother. 59:531–540.[CrossRef][Medline]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Vadose Zone Journal | |||
| Journal of Natural Resources and Life Sciences Education |
Soil Science Society of America Journal | ||||
| Journal of Plant Registrations | Journal of Environmental Quality |
The Plant Genome | |||
