Crop Science Journal of Natural Resources and Life Sciences Education
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 20 June 2006
Published in Crop Sci 46:1704-1710 (2006)
© 2006 Crop Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via ISI Web of Science (7)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Brûlé-Babel, A. L.
Right arrow Articles by Van Acker, R. C.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Brûlé-Babel, A. L.
Right arrow Articles by Van Acker, R. C.
Agricola
Right arrow Articles by Brûlé-Babel, A. L.
Right arrow Articles by Van Acker, R. C.
Related Collections
Right arrow Wheat
Right arrow Ecological Risk Assessment
Right arrow Other Models
Right arrow Crop Ecology
Right arrow Crop Genetics

REVIEW & INTERPRETATION

Modeling the Influence of Gene Flow and Selection Pressure on the Frequency of a GE Herbicide-Tolerant Trait in Non-GE Wheat and Wheat Volunteers

Anita L. Brûlé-Babel, Christian J. Willenborg*, Lyle F. Friesen and Rene C. Van Acker

Dep. of Plant Science, Univ. of Manitoba, Winnipeg, MB Canada R3T 2N2

* Corresponding author (christian_willenborg{at}umanitoba.ca)


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Different types of transgenic wheat (Triticum aestivum L.) will be ready for commercialization within the next decade, including varieties with higher yields, greater tolerance to biotic and abiotic stresses, and resistance to herbicides. The release of genetically engineered (GE) wheat may require segregation of GE and non-GE wheat to satisfy international markets. Before GE wheat is released, it is important to understand the movement of a GE trait within the agronomic production system. This study evaluated the effects of gene flow and selection pressure on the frequency of a GE trait (herbicide tolerance) in non-GE wheat and wheat volunteers. Gene flow of GE traits to non-GE wheat is inevitable through pollen or seed movement. When a GE trait does not confer a selective advantage in the production system, the frequency of the GE trait within non-GE wheat will be a function of the rate of gene flow. Low rates of gene flow will lead to low levels of the GE trait in the non-GE crop. With repeated gene flow events, the frequency of the GE trait may slowly increase in the non-GE crop. When the GE trait has a selective advantage, the frequency of the GE trait will increase rapidly in volunteer populations of the non-GE crop. Herbicide tolerance is an example of a GE trait that provides a high selective advantage when the herbicide is applied in the production system. Predictive models show that even with very low rates of initial gene flow, frequent applications of a highly effective herbicide will quickly increase the frequency of the herbicide-tolerant (HT) GE trait in volunteer populations. This has negative implications for control of volunteers and the ability to maintain tolerance levels of GE traits in non-GE wheat crops.

Abbreviations: GE, genetically engineered • HT, herbicide-tolerant


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
GENETIC ENGINEERING has provided the ability to potentially insert an unlimited number of novel traits into plants to exploit new opportunities and overcome the constraints previously imposed by reproductive incompatibility (Tolstrup et al., 2003). On a global scale, the production of GE crops is large and growing, increasing from 67.7 million ha in 2003 to 81.0 million ha in 2004 (James, 2004). Despite the unconfined release of these crops in countries such as the USA and Canada, market acceptance is still jeopardized in jurisdictions such as the EU that place restrictions on products containing GE organisms. Although GE crops are widely grown in agricultural systems worldwide, relatively little is known about the long-term effects of their introduction (Marvier, 2001). An important risk associated with these crops is transgene movement, which can result in adventitious presence of GE crop seeds in material destined for non-GE markets (Van Acker et al., 2003b). Nevertheless, transgene movement is inadequately understood (Marvier and Van Acker, 2005), particularly for intraspecific movement of transgenes within and among farming systems (Tolstrup et al., 2003).

Transgene movement from GE crops into neighboring crops, volunteer populations, related weedy species, or closely related wild relatives is a concern because it could alter farming practices, contaminate other crops, or irreversibly alter the ecosystem (Baker and Preston, 2003). Gene flow and the resulting transgene movement between plants can either be temporally mediated through volunteer populations or spatially via seed and pollen. The scale of gene flow varies widely. Gene flow via pollen generally occurs at a small scale, but rare outcrossing events can occur at great distances (Reiger et al., 2003). Gene flow by seed can, with the assistance of the seed and grain movement infrastructure, operate and influence agriculture at a global scale (Colbach et al., 2004). Furthermore, the persistence of some seeds further exacerbates gene flow and can lead to problems such as gene stacking (Hall et al., 2000). Gene stacking in this sense refers to cross-pollination resulting in genes from different GE varieties occurring together in individual volunteer plants. While gene flow is inevitable, the relative importance of gene flow to the genetic structure of populations depends on a number of factors, including population size (Farris and Mitton, 1984), duration of gene flow (Klinger and Ellstrand, 1999), distance between recipient and donor populations (Hucl and Matus-Cádiz, 2001; Manasse, 1992; Matus-Cádiz et al., 2004), and the fitness advantage, if any, conferred by the transgene (Arriola and Ellstrand, 1997). Populations which are reproductively isolated or separated beyond the pollen cloud distance are unlikely to be affected by pollen-mediated gene flow. In contrast, genes such as herbicide-resistance that confer a fitness advantage to the recipient population may have considerable impacts on population structure and dynamics (Van Acker et al., 2003a).

Development of GE wheat through recombinant DNA technologies is utilized in a number of wheat breeding programs. Herbicide-tolerant GE wheat recently had been proposed for widespread introduction but has since been withdrawn (Huygen et al., 2003). Nevertheless, several other GE traits such as disease resistance and drought resistance are expected to be incorporated into wheat cultivars in the future (Wilson et al., 2003). The traditional assumption has been that cultivar purity should be relatively simple and easy to maintain in wheat because it is predominantly a self-fertilizing species. However, because its mating system allows for partial intermating in addition to self-fertilization, off-types attributable to outcrossing are identified every year (Hucl et al., 2004; Prakash and Singhal, 2003). In wheat, pollen movement is facilitated mainly by two factors, wind and gravity. Anthers normally dehisce within the floret, followed by filament elongation and extrusion of the anthers outside of the floret (de Vries, 1971). A small amount of pollen is shed on the stigma within the floret, while 80% of the pollen is shed outside of the floret (Waines and Hegde, 2003). Florets that have not been successfully self-pollinated will remain open and receptive to pollen from other sources for up to 13 d (de Vries, 1971). Estimates of outcrossing rates in wheat are dependent on synchrony of flowering between pollen donors (males) and pollen receptors (females), the presence of receptive females, and the availability of single dominant nuclear marker genes to facilitate detection of outcrossing. Despite the primarily cleistogamous nature of wheat flowers, several studies have established that outcrossing in wheat varies among cultivars, ranging from rates as low as 0.3% to as high as 6.05% (Hucl, 1996; Hucl and Matus-Cádiz, 2001; Matus-Cádiz et al., 2004). An empirical model developed by Gustafson et al. (2005) predicted that without isolation buffers, pollen-mediated gene flow in wheat from fields > 10 ha in size would be <0.1% once harvest dilution was accounted for. Although this is well below any commercial thresholds for foreign material and labeling, even small levels of gene flow can potentially lead to rapid accumulation of transgenes under certain conditions. In fact, Waines and Hegde (2003) stated that "...there is enough evidence to show that cross-pollination [in wheat] regularly occurs and the reproductive biology of wheat is favorable to facilitate varying degrees of gene flow in a variety of situations."

Gene flow may also occur via seed movement. Seed movement may occur in a number of ways such as movement with farm or transport equipment, animals, wind, or water. Seed that has previously been contaminated with a GE trait through pollen or seed movement can contribute to the introduction of GE traits into fields that were not previously planted to a GE crop, or adjacent to a GE crop. Frequency of seed movement is expected to be highly variable and difficult to predict.

Environmental safety assessments for GE crops in North America focus primarily on the impact of the new trait on weedy characteristics of the crop, the effect of gene flow to wild relatives, the potential for the crop to become a pest, and the potential impact of the trait on nontarget organisms and biodiversity. Comparisons are made relative to a known non-GE counterpart. In most cases environmental safety assessments focus on impacts outside of the agronomic production system. However, some GE traits can have significant impacts on crop production practices, pedigreed seed purity, ability to manage volunteers, and the ability to produce non-GE wheat crops. Surveys conducted by the Canadian Wheat Board show that there is significant customer resistance to GE wheat (Canadian Wheat Board, 2003). As a result, initial release of GE wheat will require segregation of GE and non-GE wheat to satisfy different customer demands (Huygen et al., 2003; Holzman, 2001). Under these circumstances, it will be important to understand the fate of GE traits in wheat within the production system. Therefore, the objective of this study was to assess the potential effects of gene flow and selection pressure on the frequency of a GE HT trait in non-GE wheat and wheat volunteers using basic population genetics principles.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Frequency of Herbicide-Tolerant Plants Resulting from Selection Pressure
Mechanistic modeling was used to simulate the evolution of herbicide resistance in a volunteer wheat population and in non-GE wheat crops. Potential gene-flow rates in the model were calculated based on outcrossing rates reported by Hucl and Matus-Cádiz (2001). The authors compared outcrossing rates among four wheat cultivars (Katepwa, Roblin, Oslo, and Biggar) using a dominant blue aleurone trait in the pollen source to quantify outcrossing rates. Outcrossing rates varied considerably among the different cultivars (Hucl and Matus-Cádiz, 2001). Katepwa showed the lowest level (0.2%) of outcrossing and Oslo showed the highest level (3.8%) of outcrossing. Outcrossing was reported up to 27 m from the pollen source. However, it is important to note that the pollen source plot size in this study was small (5 m2) and sample sizes evaluated were low (<700 seeds sample–1). Others studies have shown that outcrossing rates in wheat fall within the range of 0.1 to 10.1% (Griffin, 1987; Martin, 1990; Hucl, 1996; Enjalbert et al., 1998; Hanson et al., 2005). Model simulations in the current study utilized outcrossing rates similar to Katepwa and Oslo to provide comparisons of low and high outcrossing rates. Although outcrossing has been reported as far as 48 m (Khan et al., 1973) and 300 m (Matus-Cádiz et al., 2004) from the pollen source, levels of outcrossing are highest in the first 10 m from the pollen source. Therefore, to simplify modeling, outcrossing and gene flow were assumed to occur within 10 m of the pollen source at a level of either 0.01 or 3%, similar to the rates observed in Katepwa and Oslo, respectively (Hucl and Matuz-Cádiz, 2001).

In the models, the evolution of resistance in non-GE and volunteer wheat populations is assumed to result from pollen-mediated gene flow initially, with resistance being conferred by a single, unlinked allele at a nuclear gene locus. Following the initial gene flow event, an outcrossing rate of 1% within the resulting population was assumed. The herbicide to which resistance is being simulated in the model is hypothetical, but provides nonselective control of both crops and weeds that lack the gene conferring tolerance to the herbicide. Resistance in this model was assumed to be inherited in a completely dominant fashion at field herbicide rates. The selection pressure used in the general selection model was set at 95% to simulate typical herbicide efficacy. The model cannot account for rare, stochastic events.

Models were developed based on mathematical equations commonly employed in population genetics to evaluate the effect of gene flow either in the presence or absence of selection pressure (Hartl and Clark, 1989). A general selection equation was modified to accommodate the primarily self-pollinating nature of wheat, and therefore the models do not assume that the populations are in Hardy-Weinberg equilibrium. The genotype frequencies in any given generation were calculated specifically for each genotype.

Frequency of homozygotes dominant for the trait (FA1A1):

Formula 1[1]

Frequency of heterozygotes (FA1A2):

Formula 2[2]

Frequency of homozygotes (FA2A2) recessive for the trait:

Formula 3[3]

where S is the frequency of selfing in the population, 1 – S is the frequency of random mating in the population, P, H, and Q are the genotypic frequencies of A1A1, A1A2, and A2A2 in the previous generation, respectively, W11, W12, and W22 are the relative fitness of the A1A1, A1A2, and A2A2 genotypes, respectively, and Formula 3 is the average fitness of the population. Formula 3 is calculated as

Formula 4[4]
Although outcrossing and gene flow were assumed to occur within 10 m of the pollen source, the frequency of the HT trait would be reduced by the dilution resulting from harvesting a large field area and therefore the effect of 0-, 10-, 50-, and 100-fold dilutions were also modeled.

Effect of Repeated Gene Flow from GE crops to Non-GE crops on Genotype Frequencies
A pollen migration model was developed to examine the effect of repeated generations of gene flow via pollen only at various rates from a migrant GE wheat population on the frequency of the HT trait in a non-GE wheat resident population. Pollen migration was modeled assuming unidirectional migration to the population under consideration (non-GE wheat crop) from the source population (GE, HT wheat crop), constant pollen migration rates and genetic composition across time in the source population, and continued planting of seed from the non-GE population (e.g., farm-saved seed). It was further assumed that resistance in this model is inherited in a completely dominant manner at field herbicide rates with resistance being conferred by a single allele at a nuclear gene locus, with the frequency of the GE trait being one in the source population. The model does not account for the effects of selection, mutation, and genetic drift nor does it account for rare, stochastic events.

When pollen migration contributes to production of progeny, the genotype frequencies produced are a function of the proportion of progeny produced as a function of the frequency of successful pollinations with the migrant population (m) and the frequency of progeny produced through mating among genotypes within the nonmigrant population (1 – m). Within the nonmigrant population, progeny may be produced through selfing (S) or through random mating within the population (1 – S).

The frequency of GE progeny produced from gene flow is a function of gene flow (m) and the frequency of the GE alleles in the pollen population mated with the GE and non-GE females from the resident population. Similarly, the frequency of GE progeny that did not result from gene flow is a function of mating within the resident population (1 – m) and the frequency of GE and non-GE alleles within that resident population. Following this, we can calculate the resulting genotypic frequencies in each generation of the resident population (non-GE crop) resulting from migration, random mating, and selfing by the following.

Frequency of homozygotes dominant for the trait (FA1A1):

Formula 5[5]

Frequency of heterozygotes (FA1A2):

Formula 6[6]

Frequency of homozygotes (FA2A2) recessive for the trait:

Formula 7[7]

where Pt+1, Ht+1, and Qt+1 are the genotypic frequencies of A1A1, A1A2, A2A2 in the next generation (t + 1); Pt, Ht, and Qt are the genotypic frequencies of A1A1, A1A2, A2A2 in the previous generation (t), respectively; Pm, Hm, and Qm are the frequencies of A1A1, A1A2, and A2A2 in the migrant population, Pm being 1 and Hm and Qm being 0, respectively, and m, 1 – m, S, and 1 – S are as described above.


   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
CONCLUSIONS
 REFERENCES
 
The fate of GE traits within production systems will depend on the selective advantage or disadvantage conferred by the trait within the production system. Many GE traits will not confer a substantive selective advantage either within or outside of the production system. For these traits, their frequency in the non-GE wheat population will be maintained at fairly low levels related to the rate of gene flow. However, GE traits that confer a selective advantage either within the production system or outside of the production system will increase in the population with each generation in which the selection pressure is present. Herbicide tolerance traits are an example of the type of trait that confers a selective advantage within the production system. The more effective the herbicide and the more frequently it is applied, the more rapid the increase in the frequency of the GE trait in the population.

Selection pressure exerts a large effect on the frequency of a GE trait in a non-GE wheat volunteer population (Fig. 1 ). In this situation, the GE trait is assumed to be an herbicide resistance trait. To simulate a field situation, gene flow rates similar to those reported by Hucl and Matus-Cádiz (2001) for Oslo (Fig. 1A) or Katepwa (Fig. 1B) were used as the upper limit. The remaining gene flow rates simulate a situation in which gene flow occurs in the first 10 m of the field, but then a larger field area is harvested that dilutes the frequency of the GE trait by 10, 50, or 100 times. In these scenarios, it is assumed that gene flow has occurred only in the initial generation. Figure 1A indicates that 50% of the volunteer wheat population will be resistant to the herbicide after only two to four generations of treatment. Even at low initial gene flow rates (Fig. 1B), 50% of the volunteer population will be resistant after only four to six generations of herbicide treatment. Therefore, even with relatively low gene flow rates, the frequency of traits that have a high selective advantage in the production system will increase rapidly with the application of the selective agent. For herbicide resistance traits, this will have a significant impact on volunteer management, crop rotation, herbicide management, and ability to maintain minimum tolerance levels of GE traits in non-GE crops. This suggests that the strict control of volunteers will be critical to the maintenance of GE-free wheat crops.


Figure 1
View larger version (26K):
[in this window]
[in a new window]
  		Fig. 1. Frequency of herbicide-resistant genetically engineered (GE) wheat volunteers following application of the herbicide. The selection pressure of the herbicide is 95%. (A) Initial gene flow rates, accounting for dilution, range from 0.003 to 3% to simulate a cultivar such as Oslo that has a high outcrossing rate. (B) Initial gene flow rates, accounting for dilution, range from 0.00001 to 0.01% to simulate a cultivar such as Katepwa that has a low outcrossing rate.

 
Control of resistant volunteer populations may be achieved through a combination of crop rotation and rotating to alternative herbicides. Crop rotation would be a very effective strategy for managing resistant volunteer wheat populations because it employs a variety of crop characteristics such as varied planting dates, growth habits, fertility requirements, and most importantly, altered herbicide requirements. Monocultures of continuous wheat, wheat–fallow–wheat–fallow, or wheat–wheat–fallow common to Northern Great Plains rotations (Gan et al., 2003; Zentner et al., 2002) would tend to magnify the problem. Moreover, rotating to alternative crops would tend to minimize the persistence of volunteer plants and also reduce the rate of outcrossing because it lessens the frequency of common borders between transgenic and nontransgenic crops. Rotating to alternative herbicides are another option, but will result in an additional cost to both adopters and nonadopters of GE wheat. For example, low disturbance direct seeding producers will be have to add an additional herbicide to their preseeding herbicide mixture to control resistant volunteer wheat (Van Acker et al., 2003a). Another option producers could use to control resistant volunteer populations is tillage. However, conservation tillage and zero-tillage producers would be reluctant to use this method due to the detrimental impact it may have on soil moisture content and soil health. Therefore, although options are available, they will likely result in increased herbicide load into the environment, additional costs to producers (particularly those practicing conservation tillage or zero-tillage), as well as creating the potential for increased tillage.

With the exception of Canada Prairie Spring wheat, current Canadian standards for pedigreed wheat seed production allow a maximum of 1 in 10000 off-types in Breeder and Select seed, and 5 in 10000 in Foundation, Registered, and Certified seed (Canadian Seed Growers' Association, 2005). Therefore, Certified seed could have a frequency of GE traits equivalent to a gene flow rate of 0.05% and still meet pedigreed seed standards. If the GE trait is resistant to an herbicide that is used frequently within the production system, it is possible that producers could rapidly increase the frequency of GE traits within their volunteer populations through their normal production practices. The rate of increase of the GE trait would be similar to the gene flow rate of 0.06% shown in Fig. 1A.

Even though gene-flow rates in wheat may be relatively low when compared with crops that are primarily cross-pollinating, the levels of gene flow are sufficiently high that it will not be possible to guarantee 0% GE trait in non-GE wheat (Fig. 1). Even with outcrossing rates as low as 0.01%, half of the volunteer wheat population will contain the HT transgene within three generations when the trait is continuously selected (Fig. 1B). In contrast, the seemingly low level of pollen-mediated gene flow (0.1%) predicted by the empirical model developed by Gustafson et al. (2005) would be expected to result in half of the volunteer wheat population containing the HT transgene within two generations according to our model (Fig. 1A). Statistically, it is neither practical nor possible to prove a tolerance level of 0% GE. Therefore, before release of GE wheat it will be important to have established standards for tolerance levels of GE traits in non-GE wheat (Van Acker et al., 2003a, 2003b, 2004). This will be important for both conventionally and organically produced crops, particularly those grown for seed. After tolerance levels are established, sampling and testing procedures can be established to guarantee that non-GE wheat crops do not exceed the tolerance levels.

When a field of GE wheat is grown adjacent to a non-GE wheat field, some outcrossing will likely occur (Gustafson et al., 2005). The level of outcrossing will depend on the synchrony of flowering between the two fields, the level of male sterility in the non-GE wheat (i.e., degree to which receptive females are available), the non-GE cultivar, distance between the crops, and wind direction. The frequency of the GE trait in the harvested seed from the non-GE crop will be influenced by the rate of outcrossing experienced and size of the field being harvested. Since the highest level of outcrossing will occur along the field margin closest to the GE crop, it is expected that the frequency of the GE trait will be greatest along the field edge of the non-GE crop and will diminish with distance from the GE crop. As the non-GE crop is harvested, it is expected that seeds containing the GE trait will be mixed with, and diluted with, non-GE wheat from the remainder of the field. Depending on how the field is harvested, the frequency of the GE trait may vary significantly from sample to sample. If the harvested grain is used for seed, the GE trait may be inadvertently introduced into even an isolated field. Similarly, wheat volunteers that remain after harvest will contain the GE trait at a frequency equivalent to the outcrossing rate. In the absence of repeated gene flow events and under situations that do not provide a selective advantage or disadvantage to the GE trait, the frequency of the trait will stabilize and remain constant within the population. If volunteer population sizes are very low, the frequency of the GE trait may increase or decrease due to random genetic drift.

When a non-GE wheat crop is grown adjacent to a GE wheat crop across multiple generations, the frequency of the GE trait is expected to increase in the non-GE wheat crop. If no other forces are acting on the population, the rate of increase will be directly related to the level of gene flow between the two crops (Table 1). If tolerance levels for a GE trait in non-GE wheat are as low as 0.25 or 0.5%, it will be difficult to maintain this standard if migration rates are relatively high (between 0.1 and 1%) (Fig. 2 ). In contrast, it should be fairly easy to maintain this tolerance under lower levels of outcrossing (0.01 and 0.001%) (Fig. 2), but even cultivars with typically low outcrossing rates could exhibit high amounts of gene flow given synchronous flowering with GE wheat, particularly with respect to the volunteer wheat population. These results stress the importance of ensuring that seed is produced under conditions that severely limit the potential for gene flow between GE and non-GE wheat. Producers that use farm-saved seed may need to rethink this strategy if there is potential for gene flow from a neighboring GE wheat crop into their non-GE wheat crop. Alternatively, they may wish to examine their harvesting and seed handling procedures to strongly reduce the potential for introduction of GE wheat into their non-GE wheat seed source and crop. Practices such as thorough cleaning of seeding, harvesting, and grain transportation equipment, as well as complete separation of GE and non-GE seed on the farm will be essential to reducing the potential for introduction of GE wheat into the non-GE seed source. It may also be necessary to temporally isolate the two production systems, such as seeding and harvesting all non-GE fields first, to ensure seeds are not carried from GE field to non-GE field.


View this table:
[in this window]
[in a new window]
  		Table 1. The effect of repeated generations of outcrossing at various rates from genetically engineered (GE) wheat to non-GE wheat on the frequency of the GE trait in a non-GE wheat population.

 

Figure 2
View larger version (19K):
[in this window]
[in a new window]
  		Fig. 2. The accumulation of a genetically engineered (GE) trait in a non-GE wheat population resulting from repeated generations of migration at rates of 0.1, 0.01, and 0.001%. Drop lines indicate the generations required to exceed potential thresholds of 0.25% (– –) and 0.5% (– –) GE wheat in non-GE wheat crops.

 
The potential for selection to increase the frequency of GE traits in volunteer populations must be considered when trying to meet standards for non-GE crops (Van Acker et al., 2003a, 2003b). Within a region such as the northern great plains of North America, where wheat forms a high proportion of common crop rotations (Van Acker et al., 2004), the volunteer wheat populations are large and stratified in space and time. In the absence of absolute prevention of seed return at harvest, absolute efficacy of control treatments and the prevention of stochastic seed return events such as those caused by hail, it is not possible to eliminate volunteers from considerations of gene flow and segregation. Figure 3 indicates that when the frequency of a GE trait within a volunteer wheat population is relatively high, even a small number of volunteers could make it difficult to meet non-GE standards. For example, if the frequency of a GE trait in a volunteer population is 50% and non-GE wheat is sown into a field at a standard seeding rate of 250 seeds m–2, as few as 6 GE wheat volunteers m–2 would lead to an adventitious presence level that exceeds 1%. Similarly, 16 and 27 volunteers m–2 would lead to adventitious presence levels that exceed 3 and 5%, respectively. Marginet (2001) reported that pretreatment wheat volunteer densities ranged from 1 to 171 plants m–2, and most frequently ranged between 20 and 40 plants m–2. Therefore, this typical density of wheat volunteers could cause concern even if the frequency of the GE trait in the volunteer population was as low as 10%. In addition, wheat volunteers continue to persist after control measures have been applied. In the most recent postcontrol survey of Manitoba, volunteer wheat was found on 15.8% of fields at an average occurrence density of 2.1 plants m–2 (Leeson et al., 2002). Furthermore, Beckie et al. (2001) presented empirical evidence of volunteer wheat in western Canada persisting for up to 5 yr on up to 9% of fields.


Figure 3
View larger version (19K):
[in this window]
[in a new window]
  		Fig. 3. Number of volunteers that would lead to 1, 3, or 5% adventitious presence of a genetically engineered (GE) trait in a non-GE wheat crop sown at 250 seeds m–2 relative to the frequency of GE plants in the volunteer wheat population.

 
In the reduced tillage farming systems typical of the Northern Great Plains region, preseed spring burn-off herbicides are currently used instead of tillage for preseed weed control (Friesen et al., 2003). Since preseed spring burndown herbicides may be applied immediately before seeding and possibly before emergence of all volunteers, producers may not be aware that there may be a high frequency of resistant GE volunteers in their fields. If the problem is not discovered before testing and marketing of the grain, this could cause economic losses to the producer as well as grain marketers. The GE wheat volunteers that are not controlled before seeding, or that cannot be controlled in crop, could also cause problems for producers of other non-GE crops such as barley (Hordeum vulgare L.) or oat (Avena sativa L.), since wheat seed is similar in size to barley or oat, and GE wheat seed is not easily removed from these crops.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
REFERENCES
 
After GE wheat is released for commercial production, there will be potential for gene flow of GE traits from GE wheat to non-GE wheat crops. This can occur through either seed movement or pollen movement. Probable customer resistance to GE traits will require segregation of GE and non-GE wheat to ensure ongoing marketability of the Canadian wheat crop (Canadian Wheat Board, 2003; Huygen et al., 2003; Wilson et al., 2003). Tolerance limits for GE traits in non-GE wheat will have to be set to meet customer demands. Since tolerance limits likely will vary among different customers, it will be important to set tolerance limits that will satisfy the majority of customers. However, it will not be possible to maintain tolerance levels of 0% GE trait. Transgene movement in wheat appears inevitable and it may be difficult to maintain coexistence between GE and non-GE wheat, particularly if there are no tolerance levels (0% tolerance) for the GE trait or if tolerance levels are very low (<1%).

The vast acreage of wheat in western Canada (10 million ha annually) suggests that some wheat fields will be grown adjacent to each other with very little distance separating them. The minimum isolation distance for production of pedigreed Breeder and Select seed is 10 m and for Foundation, Registered, and Certified seed is only 3 m (Canadian Seed Growers' Association, 2005). On the basis of the outcrossing rates and distances reported in Hucl (1996), Hucl and Matus Cádiz (2001) and Matus Cádiz et al. (2004), gene flow between GE and non-GE wheat will be a concern in a production system that requires segregation of non-GE wheat from GE wheat. In the short term there is little concern regarding gene flow of non-GE traits to GE wheat. As a result, the main focus should be on the fate of GE traits in non-GE wheat crops and volunteers.

The need to segregate GE and non-GE wheat, at least for the short-term, will require a clear understanding of the fate of the GE trait within the production system. With repeated gene flow events, GE traits that do not confer a selective advantage within the production system may increase slowly within seed populations of wheat, but may not present substantive problems for segregation of GE and non-GE crops unless tolerance levels are very low. GE traits that do confer a selective advantage within the production system are expected to increase in frequency within volunteer wheat populations. The highest rate of increase will occur for GE traits that confer a high selective advantage to an agent that is applied frequently within the production system. Resistance to glyphosate in the context of reduced-tillage farming systems is an example of such a trait. As GE traits are commercialized, it will be important to review standards for pedigreed seed production to ensure that problems are not generated for those who choose to grow non-GE crops.

Received for publication November 11, 2005.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 





This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via ISI Web of Science (7)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Brûlé-Babel, A. L.
Right arrow Articles by Van Acker, R. C.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Brûlé-Babel, A. L.
Right arrow Articles by Van Acker, R. C.
Agricola
Right arrow Articles by Brûlé-Babel, A. L.
Right arrow Articles by Van Acker, R. C.
Related Collections
Right arrow Wheat
Right arrow Ecological Risk Assessment
Right arrow Other Models
Right arrow Crop Ecology
Right arrow Crop Genetics

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