Statements and Comments

Biotechnology - Non-ASTA

ADVENTITIOUS POLLEN INTRUSION INTO HYBRID MAIZE SEED PRODUCTION FIELDS
by J.S. Burris, Ph.D.

Representing the Association of Official Seed Certifying Agencies

Funded by USDA/FAS/ICD/RSED and ASTA

Contributors: S.K. Christensen, D. Curry, A.T. Ewalt, D.S. Ireland, F.L. Kaltenberg, J.R. Keiser, M.J. Lauer, M.R. Miller, D.W. Monke, T.S. Newman, L. D. Rieffel, R.J. Sabata, D.R. Thompson, H.L. Wahls, M.T. Wall, D.O. Wilson


BACKGROUND

Beginning with the domestication of maize and especially with the introduction of hybrid seed production, seedsmen have been attempting to control maize pollination. For the plant breeder the task is simply a matter of “shoot bagging” e.g., covering both male (tassel) and female (ear-silk) flowers with paper or glycine bags. This is practical for the plant breeders nursery and for initial breeders seed
increases, however, significant seed increases require large field plantings which incorporate various practices to insure acceptable pollination control and the resulting seed purity. These measures include but are not limited to:

· Crop rotation to minimize volunteer maize plants and reduce the need for roguing.
· Selection of parent seed of high purity.
· Vigorous rouging of both male and female rows to insure only the desired parents remain.
· Aggressive detasseling of the female parent to prevent self pollination.
· Time isolation of the silking period so as not to coincide with corn in nearby fields.
· Inclusion of border rows of the pollen parent around the field to insure that the field is flooded
with the appropriate pollen and to dilute potential adventitious pollen.
· Adequate isolation distance to insure acceptable levels of protection from adventitious pollen.

In general, hybrid maize field production practices are based on seed certification guidelines and on empirical evidence gathered over the years by seed producers. These practices have remained basically unchanged for thirty or more years. Current Association Official Seed Certification Agency (AOSCA) isolation standards required for the production of hybrid maize seed (Table 1). These standards have proven to be very useful when genetic purity was defined by the morphological phenotype and when 95–98% purity was satisfactory. However, as the detection methodology improved, the applicability of the current standards has been questioned. The study reported here in was conceived to address these concerns via a thorough review of the existing literature and coordination of an industry-wide study of adventitious pollen intrusion under normal seed production conditions.

Table 1. Minimum border rows of isolation for hybrid corn production according to AOSCA Standards

Minimum Distance Field Size
From Contaminant Up to 20ac 20 ac and more
410 0 0
370 2 1
330 4 2
290 6 3
245 8 4
205 10 5
165 12 6
125 14 7
85 16 8
0 Not permitted 10


LITERATURE REVIEW

Introduction

Maize Zea mays L. is normally cross-pollinated and freely crosses with nearly all members of the genus, reported to include several hundred mutants. Most pollinations result from pollen transported by wind or gravity but there have been reports of pollination carried out by bees. There are two general situations that involve contamination by adventitious pollen. One is the occurrence of contamination in seed production fields (the focus of this investigation) and the other is the contamination of hybrid grain production fields (of primary concern to farmers).

The primary difference between the two situations relates to the amount of desirable pollen present. Depending on the planting pattern, as much as 80% of the plants in a seed field maybe either sterile or emasculated. Whereas the grain field will likely contain 100% fertile parents, which provide copious amounts of the desirable pollen to buffer against intrusion by adventitious pollen. Thus the information gathered in this study have direct application to seed production and represent a “worse case” scenario as it pertains to grain production.

In addition, this study represents a comprehensive evaluation of pollen intrusion based upon modern laboratory analyses of the collected samples by electrophoresis. However, parameters such as geography, diurnal pollen background values during the production season have not been reported but will be an important consideration during the sampling period.

Pollen Morphology and Viability
Within the grass family, maize produces one of the largest pollen grains (90-125 x 85 microns)(Smith, 1990). Maize pollen grains are mono-porate and nearly spheroidal to ovoid in shape with a slightly protruding aperature(Erdtman,1952). Pollen volume is approximately 700 x 10-9 cm3 with a weight of 250 x 10-9 g (Goss 1968). Because of its large size and even though it is disseminated by wind and gravity, maize pollen normally travels only short distances as compared to other members of this family.

Maize pollen settling velocity is at 30.95 cm . s-1 , or about an order of magnitude greater than that reported for other wind pollinated pollen species being measured (Di-Giovanni et al., 1995).

Determination of pollen viability, an important aspect of pollination potential, is problematic with some considerable range in accepted values depending on the genetics and the methods used to determine viability. It is safe to assume that pollen can maintain viability from a few hours to several days. Viability is negatively affected by elevated temperatures and reduced humidity although elevated temperature appears to result in a more rapid decline than humidity. Pollen exposed to ambient field conditions decreased to 80% viability in one hour and was 100% nonviable in two hours (Luna et al., 2001).

Pollen viability decreased more slowly under more humid conditions, to 58% after one hour, yet still fell
to zero after two hours (Luna et al., 2001). Johnson and Herrero (1981) reported that pollen viability was greatly reduced by temperatures above 38oC. While Schoper et al. (1987a ) reported considerable variability in heat tolerance associated with different genotypes. Jones and Newell (1948) reported pollen survival for up to nine days when stored in refrigerated conditions. They also reported survival of only three hours in a pollination bag under midsummer field conditions and suggested that this would be
adequate time for pollen to be transported to adjacent fields. This is in sharp contrast to the distances reported by Hutchroft (1958) when he suggested that the majority of the effective pollen fell within the first 20 feet of the individual plant.

If the distance that pollen is able to travel is not well defined, the determination of viability is even less well defined. Many authors have used simple pollen tube germination in a sucrose media while others have used the development of the extended pollen tube to determine growth into a sucrose/agar media.

In less sophisticated studies, the methods may utilize the simple occurrence of a phenotypic off-type as evidence of effective pollination. In many ways these studies are the most revealing in that they address the question of pollen viability as well as availability.

Biology of Pollination

Maize, although self-fertile and monoecious, is typically cross-pollinated by the wind because of differences in floral synchrony between male (tassel) and female (silk) flowers on a single plant. Although modern breeding efforts have tended to reduce protandry (floral synchrony) the tassel may begin to shed pollen before silks emerge. The degree of male and female floral synchrony is genotype
specific and sensitive to plant population, soil fertility and environmental stress. Usually the tassel opens completely before pollen shed begins. Pollen shed begins on the central rachis about one-third of the distance below the apex and progresses in both directions toward the apex and the base of the tassel.

Tassel branches also begin shedding somewhat below the branch apex, then extend towards the branch base and tip. The last part of the tassel to shed is typically anthers at the tip of the basal branches. Timing for this event varies with hybrid and planting date but is typically approximately 950 to 1500 GDU (GDU = Sum total of degree units, where degree units are based on average of the daily
high and low minus 50) after planting.

The typical tassel may shed pollen for 2-14 days depending on genotypic and environmental factors with the majority shed during a 5-8 day period beginning on approximately the third day after the tassel is expanded (Purseglove,1972). During the shed period the pollen is released for approximately four to five hours commencing approximately one hour after sun rise. The period may be delayed by one to
two hours if the weather is cool and cloudy. For average maturity hybrids this translates into a pollen shed period of from 6:30 to 11:00 am under normal sunny field conditions.

Each plant, depending upon genotype, is capable of producing 9,000 to 50,000 pollen grains per kernel set (Jones, 1948, Raynor, 1972). There is considerable variation in the estimates for specific tassels with a range of 14 to 50 million reported by Miller (1985) to 2 to 5 million reported in modern hybrids. If it is assumed that
approximately 1000 silks (female flowers) per plant are produced, then approximately 5,000 to 30,000 pollen grains are available for each female flower. This range is typical for wind-pollinated species.

Because of selection for female dominance the size of the average dent hybrid tassel has declined over the last three decades. A similar decline has not occurred in sweetcorn and the work of Nowakowski and Morse (1982) that reported nearly 150 pounds of pollen per acre is still reasonable. These results are based on an approximate production of 3.5g of pollen per plant and 20,000 plants per acre. Female flower production typically lags behind that of the tassel and anthers with a minimal overlap resulting in approximately a 5% self-fertilization. Silks emerge on the ear over a period of time. The spikelets at the base of the ear are the oldest and produce the longest silks which become the first to protrude from the husk. Silks originating from the middle of the ear are next to emerge followed by apical silks. The number of female spikelets per ear varies by genotype and environment but only rarely
are all the spiklets fertilized and develop into seed. The silks are receptive at emergence and can remain receptive for more than 10 days. Once fertilized the silk stops elongation and desiccates rapidly. If it is not fertilized the silk will continue to elongate until it is fertilized or cellular elongation is complete.

Pollen Dispersal

Most studies of pollen dispersal have focused on the down wind distances required to provide adequate isolation to insure production of hybrid seed with acceptable purity. These studies reported between 1940 and the late 1970s (Jones, 1948, Jones and Brooks, 1950, Hutchroft, 1958, and
Raynor et al., 1972) have used visual phenotypic characteristics to measure the dispersal. More recent
studies (Garcia et al, 1998, and unpublished sweetcorn and white corn data) have clearly shown that the majority of the pollen is deposited very close to the source. However, even though less than one percent of the pollen (Raynor et al, 1972) may be present beyond 60 meters one percent of 5 million grains per plant remains as a considerable pollen source.

Paterniani and Stort (1974) used a dominant yellow corn pollen source in the center of white cornfields of differing sizes. They demonstrated that 50% of the kernels on an individual plant could result from pollen sources within 12 meters. They also reported that as distance increases away from the pollen source, a low background level of outcrossing is attained and that remains constant at less than 0.1% out to the edge of the field.

As discussed earlier, maize pollen is one of the largest of the wind borne pollens (700 x 10-9 cm3 with a weight of 250 x 10-9 g) and generally is not widely dispersed. The degree to which it is disseminated has been investigated and a useful term to describe those results is the concept “half distance” (HD) which is the distance required to reduce the pollen amount by 50%. Considerable range in HD has been reported depending upon where and how the study was conducted. Bateman (1947) reported an HD of 3.7m while Hodgson (1948) reported a HD of 8.3m. Both of these values are consistent with the predicted values that would be expected based on the work reported by Raynor et al (1972).

In sharp contrast, Jones and Newell (1948) reported a HD of less than 25m. These differences in HD point out some of the inconsistency in data supporting the current isolation standards. This is especially important in that the reports noted relied on visual phenotypic characters and may have been even more variable if laboratory determination of genotype had been used.

The variation in the actual field measurements has prompted considerable interest in predictive models of pollen dispersal. In a study of settling velocities of various pollen types Di-Giovanni et al.,(1995) reported that maize settled nearly ten times faster than the other pollen types. These rates were measured using the methods described by McCubbin (1918). Other workers have used other models with varying degrees of success.

One problem is that because of its size maize pollen does not behave in a Gaussian fashion. Gaussian models require that the particles are distributed normally around their source. The vertical distribution of maize pollen normally does not increase with distance from the source and typically drops off rapidly (Raynor etal. 1972). Even though considerable effort has been devoted to attempting to model maize pollen distribution, model accuracy has done little to refute the actual sampling results. Models provide good descriptive data that can be helpful in explaining specific field results. But, since many important variables, i.e., wind speed and direction and surface turbulence, cannot be predicted, they do not provide strong predictive data.

In a study on outcrossing, Jones and Brooks (1950) discussed the effect of distance and border rows as production practices in Oklahoma. They reported that isolation distances of greater than 40 rods (200m) from the potential contaminate was required to result in less than 2% contamination. Jones and Newell (1948) observed < 3 % outcrossing at 250m, but Airy (1955) found little difference in outcrossing between 100 and 200m from the potential contaminate. The effect of border rows is inconclusive and dependent on the direction of the contaminate. But in general when isolation distances
were less than 200m five border rows were required to achieve satisfactory purity (Jones and Brooks 1950).

Although few authors have addressed the contribution of border rows. Other outcross data include those of Jones and Newell (1948), who reported 8.9 % at 125m and less than 3 % at 200 meters. Airy (1955) found little difference in outcrossing between 100 and 200m from the potential contaminate. The industry standards that were developed using visual characters have survived relatively unchanged for more than 50 years. With the additional concerns raised by the relatively recent inclusion of modified traits, an examination of these standards is in order.

Methods and Materials

The industry-wide study of seed production fields was conducted during the 1998, 1999, and 2000 production years. Production locations across the corn belt were asked to nominate production fields for inclusion in this study. Potential fields had to meet the following criteria: satisfactory male and female stands, appropriate nick (reproductive match) between the seed and pollen parents, and pollen shed in the potential contaminate should match the seed field silk emergence.

In the selected fields, additional agronomic data were collected, including: previous crop, dryland or irrigated, male pollen shed rating (1=good, 2=fair, 3=poor), row pattern (female:male), field size, block size, number of border rows, plant population of male and border rows, compass orientation to contaminate field, distance to contaminate, and size and type (CC=conventional corn, WC=white corn, SC=sweet corn, PC=popcorn or purple corn) of contaminate field.

Prior to harvest, fields were sampled along a line perpendicular to the source of the contaminate and in
the pattern described in Figure 1. Each field map had an arrow denoting a "sampling path." Each path had five sampling locations. These locations were within the following female blocks; 1st, 3rd, 6th, 10th and the middle of the field or 200m (whichever was closest to the potential-contaminant). At each of these locations a sample was collected from the center two rows of the designated female block. In the center two rows, a random plant was selected and then all ears with seed from the next ten consecutive plants were collected. Then ten consecutive plants from the other center row were collected moving in
the same direction (Figure 2). If the female blocks were not aligned across the potential-contaminant (i.e., rows run away from the contaminant), the center two rows of the female block that corresponded to the middle of the potential contaminant were sampled at distances from the most inside male border row: 2m, 10m, 20m, 35m and the mid-point or 200m, whichever was closest to the potential contaminant.

Figure 1. Sampling path indicated by red arrow. The black points on the red arrow denote the five sampling positions in the 1st, 3rd, 6th, 10th female blocks and the middle female block or 200m (whichever was closer to the potential contaminant).

Figure 2. Ear collection points within selected female blocks. All ears that set seed from ten consecutive plants were collected from each of the middle two rows. Each sample from each position contained all ears that set seed from 20 consecutive plants; ten plants from each of the center two female rows.

After collection, the samples were husked and dried to 12% moisture using standard techniques. After drying, samples were shelled and screened (scalped) to pass a 26/64 and held over a 16/64 screen. The composite samples were submitted to electophoresis testing using the standard number of loci appropriate for the genotype involved. Entire seed fields were harvested, dried and conditioned as a lot. Weighted outcross values for each field were calculated from the five samples collected to determine whole field estimates of outcrossing. Sample results were weighted rather than averaged because each sample represented a larger proportion of the field as sampling progressed from the edge to the center of the field, away from the adventitious source. Sample results were also weighted based on the size and shape of the seed field and on the proportion to the total seed field area represented by each electrophretic sample.

RESULTS AND DISCUSSION

Because of the nature of this study, a level of caution needs to be used when interpreting or applying the results. Most agricultural scientists are familiar with experimental methodology which controls or provides measurement of the known variables. The work reported here is not such an experiment and, in fact, is more similar to studies typically reported by the social scientists or economists. Validity or accuracy of the information is not in question, rather the typical tests of significance are based on
important assumptions related to the values being representative of the sampled population and that the
values are distributed normally. The sampling design is statistically sound, however, as is clear from the field selection criteria, the fields sampled would likely be considered problematic by most producers and typically represent only a small fraction of the total number of production fields. Also because the outcross percentages are in general less than 5% and in most cases 0%, the distribution is skewed towards the bottom end of the scale with many of the values being zero. Thus the variance associated with these numbers is distributed differently than would normally be encountered in a typical randomized
block design. This is not to diminish the accuracy of the data but to point out that it has what is referred
to as “issues of colinearity”, which is to say that many important variables could not be controlled and in
the case of wind direction and velocity, were not measured nor were all out cross determinations made by the same laboratory although in general the same criteria were used.

Accepting these concerns, the data represents an extremely ambitious and expensive study which thanks to the cooperation of the hybrid seed corn industry, provides the most robust data available in the last forty years that addresses the issue of adventitious pollen movement in actual seed production fields. In most tables, two outcross
values will be reported, they are Margin (the weighted average of the first four sampling locations) and
background (the value from the midpoint or 200 meters into the field). The background value is very similar to the weighted average for the entire field.

Independent Variables

Year of sampling is one of the major variables having a significant effect on the outcross percentages (Table 2). There are a number of environmental and agronomic factors, which contribute to this variation but since they are nearly impossible to control or define the year effect will be averaged into most of the additional variables presented.

Table 2. The effect of year on the percentage of outcross occurrence in hybrid seed production fields.

Production
Year
Number
of Fields
Mean OC
Margin
Mean OC
Background
Standard
Error Mean
1998 60 0.98 0.73 0.12
1999 94 2.05 1.12 0.22
2000 212 2.03 1.27 0.15

Field location has a major impact on the level of adventitious pollen intrusion, Figure 3. It is also clear
that the level of intrusion is quite random and attempts to improve purity by moving production to any specific location is unlikely to be successful. Those factors which contribute to adventitious intrusion are clearly not related to any specific location in the central Midwest.
Sample location effect on outcross percentage demonstrates a classical pattern of protection near the border rows (Table 3.). There is some apparent protection provided to the first sample location by the border rows, then an increase in outcross percentage initially with increased distance into the field, then a gradual decline as the distance into the field increases ultimately reaching its lowest value (Background) at 200m or the middle of the field. Just as importantly, the median value remains at 1.00
until the middle of the field where with a median of 0.00 more than one-half of the values are without contamination. Further the values at 200m values very nearly describe the weighted average value for the entire field, suggesting that high purity seed can be produced even under the most difficult conditions.

Figure 3. Background outcross percentage by location across year.

Table 3. The effect of sample location on the percentage of outcross occurrence in hybrid seed production fields averaged across years.

Sample Location
Meters from
Border Row
N Mean
Out Cross
Percentage
Standard
Error
Mean
Median
2 353 1.98 0.16 1.00
10 361 2.00 0.15 1.00
20 362 1.81 0.15 1.00
35 364 1.72 0.14 1.00
200 351 1.11 0.09 0.00

The year effect is graphically represented in Figure 3. The patterns are similar in two of the years and differ somewhat from the third year. However, data from all three years indicate that the potential for contamination decreases rapidly reaching a low level within 20 to 30 meters in from the edge of the field. The variation among years clearly demonstrates the difficulties faced by production departments when they attempt to insure adequate isolation. But it is also clear that the practices currently in place are capable of producing seed of exceptionally high purity. This is especially true considering that the detection methodology used in this study was a magnitude more sophisticated than the morphological traits used as measures in earlier studies.

Effect of Field Location on Outcross Percentage
0
0.5
1
1.5
2
2.5
3
3.5
4
0 50 100 150 200 250
Meters from Border Row
Outcross Percentage
1998
1999
2000

Figure 3. Outcross percentage effected by sample location and year of production.

The Contaminate Direction has often been considered an important factor in mediating the severity of the outcross contamination (Table 4.) The foundation for that position is supported by the mean values from the first four sample locations. Sources to the north and south are similar in pollen pressure and somewhat more severe than sources to the east and as expected sources to the west are the most
severe. However, when the outcross levels from the center of the field are examined there is little if any pattern. This suggests that the adventitious pollen in the center of the field or Background is not originating in any particular field or coming from any specific direction.

This poses a significant challenge for all seed producers operating in the central cornbelt. This is not a new or unexpected revelation but likely reflects the prevailing conditions under which seed corn is produced. It does underscore the importance of achieving a very good match (nick) between the male and female flowering.

Table 4. The effect of contaminate direction on the percentage of outcross occurrence in hybrid seed production fields averaged across years.

Contaminate
Direction
Field
Number
Mean OC
Margin
Mean OC
Background
Standard
Error
Mean
North 62 1.85 1.14 0.29
South 96 1.81 1.01 0.22
East 89 1.64 1.20 0.23
West 105 2.13 1.11 0.23

Table 5. The effect of male classification on the percentage of outcross occurrence in hybrid seed production fields averaged across years.

Male Class Number
Of Fields
Mean OC
Margin
Mean OC
Background
Standard
Error
Mean
Good 146 1.73 0.98 0.15
Fair 163 1.81 1.02 0.14
Poor 57 2.37 1.94 0.35

Male Classification is routinely considered by seed producers and often effects the field planting plans. A prolific male which is well matched with the female parent is often considered the best insurance against adventitious pollen. To compensate for poor male performance, the production plans will often call for increased border rows and increased male percentages. Assuming that these
precautions were taken in the fields represented in this study, one would expect to see little, if any, influence resulting from male classification. The results presented in Table 5 show that the quality of the male parent strongly influences the level of outcrossing present. This effect is consistent both along the margin of the field and in the center. This obvious yet often-overlooked trait may be an important tool to improve the genetic purity of the seed produced. However, the prevalence of poor male
performance is likely strongly linked to the shifts in harvest index, which result in higher grain yields. Thus, modifying this trait will not be easily accomplished in high yielding genotypes.

When production plans require a poor male a potential solution often implemented is an increase in the male percentage (planting pattern) (Table 6). Although this would seem to be a reasonable solution the data do not support this alternative. There is little or no improvement in out cross contamination by modifying the male percentage. The confounding influence of male classification and other agronomic characteristics may make a clear identification of this Table 6. The effect of male percentage on the percentage of outcross occurrence in hybrid seed production fields averaged across years.

Male Percentage Number
Of Fields
Mean OC
Margin
Mean OC
Background
Standard
Error
Mean
17% 6 2.17 2.33 1.56
21% 109 1.52 0.88 0.15
25% 241 2.01 1.25 0.13
33% 4 2.28 0.26 0.49
50% 4 2.56 1.75 1.52

Management option difficult from the data available in this study. Seed field size is a management option which is often considered when planning for high quality seed
production. The results in Table 7 indicate that in general there are two size categories which appear to impact the final quality. Those fields with less than 100 acres consistently resulted in greater pollen intrusion regardless of whether the margin or the center of the field was sampled. Parent seed fields are often small, thus the improved purity resulting from large field size will be difficult to achieve.

Recognizing this limitation will force parent seed producers to manipulate other management variables to
compensate for the reduced field size and still achieve a high degree of purity.

Table 7. The effect of seed field size on the percentage of outcross occurrence in hybrid seed production fields averaged across years.

Seed Field Size ac Number
Of Fields
Mean OC
Margin
Mean OC
Background
Standard
Error
Mean
1 – 60 ac 53 2.22 1.19 0.28
61 – 80 ac 74 2.30 1.41 0.28
81 – 100 ac 45 2.30 1.51 0.34
101 – 120 ac 53 1.53 0.94 0.27
121 – 140 ac 54 1.50 0.94 0.21
141 - 180 ac 48 1.42 0.69 0.21
> 180 ac 39 1.57 1.29 0.26

The contribution of distance to improving seed purity is much less clear than was expected (Table 8).

When isolation distances are compared over a large number of fields there is a clear advantage to increasing the isolation distance. However, the confounding effects of wind direction and other uncontrolled variables may over shadow the contribution of distance.

Table 8. The effect of distance to the contaminate source on the percentage of outcross occurrence in hybrid seed production fields averaged across years.

Distance to
Contaminate
Number
Of Fields
Mean OC
Margin
Mean OC
Background
Standard
Error
Mean
1 – 50m 119 1.84 0.92 0.16
51 – 75m 67 2.79 1.79 0.31
76 – 125m 163 1.62 1.11 0.15
> 125m 13 0.73 0.55 0.21

Outcross percentages are reduced when the isolation distance is increased from 1-50 meters to 76–125 meters. Not unsurprising this trend was less evident in the middle of the field, which would be confounded by the distance to the center of the field.

The relationship between field size and isolation distance is shown in Figure 4. When the field size is relatively small (< 100ac), there is little difference between the fields at distances of 125m or less which as a group exhibit higher values than the fields at distances greater than 125m. As field size increases to greater than 140ac, there is little change in contamination level associated with increased distance from the contaminate source. Allowing that the number of fields represented at each distance is not equal and with the associated variation, the data do not show a strong relationship between increased
distance and reduced contamination level.

Field Size and Distance
0
1
2
3
4
5
6
1-60ac
61-80ac
81-100ac
101-120ac
121-140ac
141-180ac
>181ac
Field Size acres
Outcross Percentage
1-50m
51-75m
76-125m
>125m

Figure 4. Outcross percentages as effected by field size and isolation distances averaged across year.

Border rows are often inserted to insure that the edge of the production field adjacent to a potential contaminate will be flooded with the desirable pollen (Figure 5). Increasing the number of border rows does not appear to have a significant effect on the outcross percentage measured either just adjacent to the border rows or in the center of the field. Again this is a set of data which is confounded by the adjustments related to male classification and may not be representative of values found when adjustments for male classification are not made.

Border Row Effect
0
0.5
1
1.5
2
2.5
3
3.5

Field Margin Background
Field Location
Outcross Percentage
6 Rows
12 Rows
14 Rows
16 Rows
24 Rows

Figure 5. Effect of border row number on the outcross percentage in the field adjacent to the border rows and in the middle of the field.


INTERPRETIVE SUMMARY

This study was a departure from the traditional balanced randomized block design experiments that are common in the agronomic literature. It does, however, represent the largest published study of its kind under actual seed production conditions. Conducted over a three year period, it included wide participation by the seed corn industry bringing diverse environmental and genotypic differences to the data set. Before discussing the variables associated with pollen intrusion, it is important to acknowledge that the outcrosses identified could have originated as minor contaminates in the parental seed, or from unwanted volunteer plants that were not removed.

It is clear from the results of the outcross percentages by sample location that the pollen cloud associated with a specific field extends some meters beyond the field edge. It is also clear that the dominance of that cloud increases rapidly as the distance into the field (protective cloud) increases and the influence of most measurable factors, except male quality, rapidly approaches insignificance. The
data suggests that border row number is not important, but the true relationship is likely masked because of the tendency of border row number to go up when male quality goes down. However, because male quality is such a significant factor, the increased border male population has little impact.

Distance to the contaminate source is important but its contribution to reducing adventitious pollen intrusion is often over shadowed by other factors such as wind intensity, direction and the protective strength of the field pollen cloud. All this considered, this study clearly demonstrates that exceptionally high quality seed can be produced in the central cornbelt when reasonable precautions are implemented.


ACKNOWLEDGEMENTS

The results presented in this study represent the work of hundreds of individuals who participated from the planning through to sampling and to the laboratory personnel who conducted the testing. These individuals contributed thousands of hours of labor which provided the data that has been and continues to be analyzed. The companies that participated include:

Burrus Brothers & Assoc. Growers
Cargill Inc.
Curry Seed Co. Inc.
Dairyland Seed Co. Inc.
Fontanelle Hybrids
Garst Seed Co.
Golden Harvest, JC Robinson
Great Lakes Hybrids, Inc.
Kaltenberg Seed Farms
LG Seeds
Monsanto Seeds Co
Mycogen Seeds
Pioneer Hi-Bred Int. Inc.
Syngenta Seeds Inc.
Wyffels Hybrids, Inc.

Special recognition is due Dale Ireland, Dale Wilson and Michael Lauer for personally devoting countless hours to assembling this data set, contributing to the analysis and reviewing this manuscript.


REFERENCES

Airy JM, 1955. Production of hybrid corn seed. In: Corn and corn improvement, chapter IX: 379-422.

Bateman AJ, 1947. Contamination of seed crops. II. Wind pollination. Heredity I: 235-46

Di-Giovanni, F. P.G. Kevan and M.E. Nasr. 1995. The variability in settling velocities of some pollen and spores. Grana 34:39-44.

Erdtman G. 1952. Pollen Morphology and Plant Taxonomy. Angiosperms. Almquist and Wiksells, Sweden

Garcia, C., M.J. Figueroa, M.R. Gomez, L.R. Townsend, and J. Schoper. 1998. Pollen control during transgenic hybrid maize development in Mexico. Crop Sci. 38:1597-1602.

Goss JA. 1968. Development, physiology and bio-chemistry of corn and wheat pollen. Bot. Rev. 34: 333-358

Hodgson HJ. 1949. Flowering habits and pollen dispersal in Pensacola Bahia grass, Paspalum natatum. Flugge. Agron. J. 41: 337-343

Hutchcroft, C.D. 1958. Contamination in seed fields of corn resulting from incomplete detasseling. Agron J. 267-271.

Johnson, R.R. and M.P. Herrero. 1991. Corn pollination under moisture and high temperature stress.

Proc. Corn Sorghum Res. Conf. American Seed Trade Assoc. 36:66-77.

Jones MD and Newell LC. 1948. Longevity of pollen and stigmas of grasses: Buffalograss, Buchloe dactyloedees (NUTT) Engelm., and corn, Zea mays L. J. Am. Soc. Agron. 40 (3): 195-204

Jones MD & Brooks JS. 1950. Effectiveness of distance and border rows in preventing outcrossing in corn.

Oklahoma Agricultural Experimental Station Technical bulletin No 38
Luna, S., V.J. Figueroa, M.B. Baltazar, M.R. Gomez, L.R. Townsend, and J.B. Schoper. 2001.

Maize pollen longevity and distance isolation requirements for effective pollen control. Crop Sci. 41:1551-1557.

McCubbin, W.A. 1918. Dispersal distance of urediniospores of Cronartium ribicola as indicated by their rate of fall in still air. Phytopathological Notes 8:35-36.

Miller PD, 1985. Maize Pollen: Collection and Enzymology. Chapter 45, pp 279-282.

In: Sheridan WF (ed.). 1985. Maize for Biological Research. A Special Publication of the Plant Molecular Biology
Association, USA

Nowakowski J and Morse R. 1982. The behaviour of honey bees in sweet corn fields in New York state. American Bee Journal, January: 13-16.

Paterniani E and Stort AC. 1974. Effective maize pollen dispersal in the field. Euphytica, 23: 129 – 134
Purseglove JW. 1972. Tropical Crops. Monocotyledons 1. Longman Group, London

Raynor GS, Ogden EC, and Hayes JV. 1972. Dispersion and Deposition of Corn Pollen from Experimental Sources. Agron. J.64: 420–427

Schoper, J.B. R.J. Lambert, B.L. Vasilas and M.E. Westgate. 1987a. Plant factors controlling seed set in maize: the influence of silk pollen, and pollen, and ear-leaf water status and tassel heat treatment at pollination. Plant Physiol. 83:121-125.

Schoper, J.B. R.J. Lambert, and B.L. Vasilas 1987b. Pollen viability pollen shedding, and combining ability for tassel heat tolerance in maize. Crop Sci. 27:27-31.

Smith EG. 1990. Sampling and identifying allergenic pollens and molds. Blewstone Press, Texas.