The effect of flowering time and distance between pollen source and recipient on maize

Abstract

Field experiments were conducted in Central Taiwan for 2 crop seasons to examine the effect of non-coincidence flowering on the cross-pollination (CP) rate of maize at various distances. Four local maize hybrid varieties with different flowering dates and one local maize variety were sown as the pollen sources and recipient, respectively. All varieties were sown on the same day to simulate the real situation of coexistence in which adjacent fields are sown with different genetically modified (GM) and non-GM varieties of maize. The CP rate was <0.2% at a distance of 3 m for the first crop season when the flowering time for the recipient was 5 d later than that of the pollen source variety. The CP rate was <0.02% at all distances for the second season when the flowering time for the recipient was 7 d later than that of the pollen source variety. The CP rate was <1% at a distance of 0.75 m when the flowering time was 3 d later. However, varieties with closer synchrony may result in a CP rate of >1% at a distance of 1.5 m and <1% at 2.25 m. Temporal separation and isolation distances can work together in Taiwan with fragmented landscapes to minimize the adventitious presence of one crop with another.

Abbreviations

AVK

the average total kernel number per ear in the whole field

CP

cross-pollination

GM

genetically modified

Introduction

The increasing commercial cultivation of genetically modified (GM) crops worldwide encourages the establishment of coexistence strategies for GM, conventional, or organic cultures. The pollen-mediated gene flow from GM crops can result in the adventitious presence of GM material in conventional crop produces. During the past 2 decades, biotechnology studies were boosted to develop GM organisms in Taiwan. No GM crops are currently planted in the open fields of Taiwan. All GM foods in the domestic market are from foreign imports, particularly maize and soybean.

Maize (Zea mays, L) is a monoecious, diclinous, protandrous, and wind-pollinated plant with abundant pollen. An estimated 2 to 5 million pollen grains are produced by a maize plant.¹ Pollen shedding may last for 5 or 6 d, and the total pollen shedding period may last 14 d because of the low level of early and late pollen release.² The pollen shedding of maize occurs before the silks are receptive because of its protandrous characteristic, but at most 5% self-pollination can occur when pollen shedding and silking overlaps.³ Pollen-mediated gene flow may result from the adventitious mixing of non-GM maize from neighboring GM fields.^4-6

Gene flow can be monitored by measuring a recipient field’s levels of cross-pollination (CP) at various distances from pollen source. The CP rate of a maize recipient is defined as the percentage of off-types in a progeny by detecting the xenia effect in the progeny. The main factors that determine the CP rate in maize are the synchronicity of flowering, and distance between pollen source and recipient.⁵

Numerous studies regarding estimating the isolation distances to meet different thresholds have been conducted recently.^4,7-15 These studies indicated that a range of isolation distances between 20 to 25 m is recommended for maintaining the CP rate below the tolerance threshold of 0.9%. However, an isolation of at least 50 m is necessary for recipient fields that are smaller than 0.5 ha or long and narrow fields.⁵

The majority of these field trials involved planting GM and non-GM maize varieties in an adjacent field and sowing both varieties on the same day to increase the synchrony of flowering, which assisted in evaluating the CP in a worst-case scenario for the coexistence of GM and non-GM maize plants. A closer synchrony between a pollen source and recipient may increase the CP rate. Silks are receptive for approximately 5 d after silking begins and senescent within 8 d when not fertilized.^16-18 Successful fertilization requires close synchrony between the receptive silks on the female flower and pollen shed by the male flower.²

The effects of asynchrony of flowering time between source plants and recipients on the CP rates have been studied. Halsey et al.¹⁰ conducted the first practical experiments in California and Washington to evaluate the relationship of distance and temporal separation on pollen-mediated gene flow for achieving genetic purity in maize. They concluded that increasing the temporal separation may reduce the isolation distance. Outcrossing was <0.01% at 500 m when the source and recipient flowered at the same time, whereas this level was achieved at 62 m or less when 2-wk of temporal separation was used. Della Porta et al.¹³ used the pollen source planted within a recipient field of maize hybrids that had different flowering time. They showed that if there were only 3 d of difference in flowering time between the pollen source and recipient, little or no reduction of the CP rate was observed. However, 25% and 50% reductions of the CP rate were apparent for 4–5 d and 6 d of difference, respectively. Palaudelmás et al.⁵ showed that improving the coexistence through temporal separation based on appropriate delays in sowing dates is achievable.

The purpose of this study was to detect the effect of non-coincidence flowering on the CP rate of maize at various distances in Taiwan. Four local maize hybrid varieties with different flowering periods and one maize variety were chosen as the pollen sources and recipient, respectively. All varieties were sown on the same date to simulate an actual situation of coexistence in which adjacent fields of GM and non-GM maize are sown with different varieties. Moreover, we examined the combining effect of time and distance together to minimize the extent of the CP rate. This study would provide an experimental reference in Taiwan and elsewhere with similar ecosystems.

Results

Table 1 shows the synchrony between the pollen shed from all pollen donors and silking of the recipient for the 2 crop seasons. The source plots in the first season reached 50% pollen shed by May 12 (Honey 236), 14 (Black Pearl), 19 (Tainung No. 1), and 26 (Tainung No. 4), 2010. In the first season, the recipient field reached 50% pollen shed by May 14 and 50% silking by May 15. The differences in flowering dates between the pollen donors and recipient were 2 d early, and 0, 5, and 12 d late, respectively. Temporal separations were expressed as +2, 0, −5, and −12 based on the flowering date of the recipient. The source plot in the second season reached 50% pollen shed by November 20 (Honey 236), 25 (Black Pearl), 29 (Tainung No. 1), and December 2 (Tainung No. 4), 2010. In the second season, the recipient field reached 50% pollen shed by November 22 and 50% silking on November 24, 2010. The differences in flowering dates between the pollen donors and recipient were 2 d early, and 3, 7, and 10 d late, respectively. Time separations were expressed as +2, −3, −7, and −10.

Table 1.

Experimental sites in 2010 planting material, dimensions, sowing date, tasseling date and silking date

Cropseasons	Dimensions	Size of source plot	Hybrid name	Sowing date (planting)	Flowering date (50% pollen shed)	Silking date (50 % silking)
1st	50 × 100		Snow Jean (white wrinkled) (Pollen recipient)	15 Mar	14 May	15 May
		15 × 20	Honey 236 (yellow wrinkled) (Pollen source)	15 Mar	12 May
		15 × 20	Black Pearl (purple round) (Pollen source)	15 Mar	14 May
		15 × 20	Tainung 1 (yellow round) (Pollen source)	15 Mar	19 May
		15 × 20	Tainung 4 (white round) (Pollen source)	15 Mar	26 May
2nd	50 × 100		Snow Jean (white wrinkled) (Pollen recipient)	29 Sep	22 Nov	24 Nov
		15 × 20	Honey 236 (yellow wrinkled) (Pollen source)	29 Sep	20 Nov
		15 × 20	Black Pearl (purple round) (Pollen source)	29 Sep	25 Nov
		15 × 20	Tainung 1 (yellow round) (Pollen source)	29 Sep	29 Nov
		15 × 20	Tainung 4 (white round) (Pollen source)	29 Sep	2 Dec

The prevailing wind was generally presumed to originate from the south direction during the flowering periods of the first season. However winds most commonly originated from the NW direction and from the NNW, NW, W, or WNW direction (Fig. 1a) with an average daily wind speed of 2 to 4 m s⁻¹ (Fig. 2a). In the second season, winds originated from the N and NNW (Fig. 1b) with an average daily wind speed of 2 to 5 m s⁻¹ during the second season (Fig. 2b). This weather pattern was consistent with common wind patterns in the second season. The average wind speed was higher (3.21 ± 0.62 m s⁻¹) in the first season than in the second season (2.78 ± 0.91 m s⁻¹). Conversely, the maximum wind speed was 11.2 m s⁻¹ in the first season and 15.3 m s⁻¹ in the second season, respectively.

Figure 1.

Frequencies of gust wind directions measured hourly during pollen shed at 2 Taiwan agricultural research institute in 2010. The daily wind 3 measurements from 6 a.m. to 4 p.m. (a) the first crop season; (b) the second 4 crop season. 5.

Figure 2.

Mean hourly wind speed (m s-1) during pollen shed at Taiwan agricultural 6 research institute in 2010. The daily wind measurements from 6 a.m. to 4 7 p.m. (a) the first crop season; (b) the second crop season. 8.

The results of the weather pattern indicated that the recipient maize in the south direction was the downwind for both seasons (Fig. 3). The CP that resulted from the different pollen sources was calculated by summing the CP rates from the 4 pollen sources to achieve the overall CP rate, detecting the influence of wind direction to the pollen flow Table 2 shows the average overall CP rates in the upwind and downwind directions at all distances for the first season. The maximum overall CP rate occurred in the first row of the recipient field next to the source plot. The average overall CP rate at 0.75 m was 1.63% downwind and 0.27% in the upwind direction. The average overall CP rates at the same distance from the pollen source were higher in the downwind than the upwind, indicating the effect of wind direction during the flowering period. The extent of the overall CP rate in the subsequent rows declined rapidly from the source plot. Significantly low levels (0%–0.24%) of the overall CP rate were recorded at 27 m for the upwind direction and reached 0% at 12 m. The average overall CP rate was 0.19% at 3 m for the downwind direction.

Figure 3.

Field design of the experiments in 2010. (a) the first crop season (b) the 9 second crop season. The dark area indicates pollen sources and the rest is 10 recipient maize. The recipient in the south direction is downwind for both 11 seasons. 12.

Table 2.

The mean and range of total CP rate (%) in the first crop season

	Upwind (north)			Downwind (south)
Distance (m)	Range (%)	Mean ± sd	Distance (m)	Range (%)	Mean ± sd
0.75	0–2.72	0.27 ± 0.63	0.75	0–8.70	1.63 ± 2.33
1.5	0–0.54	0.05 ± 0.14	1.5	0–8.82	1.35 ± 2.44
2.25	0–0.54	0.04 ± 0.13	2.25	0–4.35	0.60 ± 1.19
3	0–0.36	0.03 ± 0.08	3	0–1.09	0.19 ± 0.26
6	0–0.72	0.04 ± 0.16
9	0–0.72	0.14 ± 0.25
12	0–0	0.00 ± 0.00
15	0–0.54	0.02 ± 0.11
21	0–0.12	0.01 ± 0.03
27	0–0.24	0.02 ± 0.02

The average overall CP rate in the second season declined when the distances from the source plot increased with the exception of the second row (1.5 m) in the downwind direction (Table 3). The average overall CP rate for the upwind direction at 0.75 m was 0.53% and decreased in the subsequent rows. Regardless of the upwind or downwind directions, the average overall CP rate was higher in the second season than in the first season (Tables 2 and 3). The variations of the overall CP rate were large in both crop seasons.

Table 3.

The mean and range of total CP rate (%) in the second crop season

	Upwind (north)			Downwind (south)
Distance (m)	Range (%)	Mean ± sd	Distance (m)	Range (%)	Mean ± sd
0.75	0–5.89	0.53 ± 1.26	0.75	0–12.12	2.85 ± 3.87
1.5	0–5.11	0.42 ± 1.06	1.5	0–14.63	3.26 ± 4.66
2.25	0–1.90	0.37 ± 0.56	2.25	0–4.68	0.66 ± 1.05
3	0–1.30	0.24 ± 0.33	3	0–1.30	0.22 ± 0.36
			6	0–1.56	0.18 ± 0.36
			9	0–2.25	0.26 ± 0.47
			12	0–0.95	0.13 ± 0.22
			15.75	0–0.61	0.15 ± 0.18
			21.75	0–0.52	0.05 ± 0.12
			27	0–0.52	0.06 ± 0.13

The effect of the temporal separation on gene flow was examined by planting 4 maize varieties with different phenotypes and flowering dates as pollen donors. The pollen recipient and donors were sown on the same date. The cultivation record shows that the flowering periods of the 4 pollen donors were approximately 1 week early, synchronous, 1 week late, and 2 weeks late compared to the pollen recipient. As shown in Table 4, the best synchrony was obtained from the Black Pearl followed by the Honey 236 for the first season. At 0.75 m in the downwind direction, the average CP rate from the Black Pearl (1.0735%) was higher than the other pollen donors (Table 4). The average CP rate declined to 0% at distance of 1.5 m from the Tainung No. 1, but Tainung No. 4 was the worst synchronous variety and never reached the average CP of 0% within 3 m. The maximum wind speed between the 3 d before and after the flowering date for Tainung No. 1(11.2 m s⁻¹ between May 16 and 22) was lower than Tainung No. 4 (16.3 m s⁻¹ between May 23 and 29). Furthermore, uneven plant growths of the Snow Jean resulted in a partial overlaying of the flowing period with Tainung No. 4. As a consequence, this led to a lower average CP rate from Tainung No. 1 compared to Tainung No. 4.

Table 4.

Effect of distance and temporal separation on gene flow in the first season

Variety	Day	Flowering Date	Distance (m)	No. of observed kernels	Total no. of ears	Observed CP rate (%)
Snow Jean	0	14 May
Honey 236	+2	12 May	0.75	53	54	0.3556
			1.5	40	61	0.2376
			2.25	29	70	0.1501
			3	25	59	0.1535
Black Pearl	0	14 May	0.75	160	54	1.0735
			1.5	170	61	1.0097
			2.25	70	70	0.3623
			3	8	59	0.0491
Tainung 1	−5	19 May	0.75	11	54	0.0738
			1.5	0	61	0.0000
			2.25	0	70	0.0000
			3	0	59	0.0000
Tainung 4	−12	26 May	0.75	29	54	0.1946
			1.5	23	61	0.1366
			2.25	12	70	0.0621
			3	4	59	0.0246

The best synchrony was obtained from the Honey 236 in the second season (Table 5). The flowering date of the Black Pearl was 3 d later than that of the Snow Jean, but the flowering period was synchronous in the first season. The delay in flowering was probably caused by the meteorological differences between the 2 crop seasons. During the first season, temperatures changed from low to high and the day length from short to long. However, these changes were reversed for the second season. For the downwind direction at 0.75 m, the average CP rate from the Honey 236 (2.0746%) was higher than the other varieties of pollen sources (Table 5). No CP was detected at 17.25, 17.25, and 2.25 m from the Black Pearl, Tainung No.1, and Tainung No. 4. However, the average 0.0296% CP rate from the Honey 236 was still traceable at 23.25 m, which indicated that the synchronization of the pollen sources' and recipient’s flowering periods influenced the extent of CP. The average CP rates at the same distance from the pollen source were higher in the downwind than the upwind for both crop seasons (Tables 6 and 7).

Table 5.

Effect of distance and temporal separation on gene flow in the second season

Variety	Day	Flowering Date	Distance (m)	No. of observed kernels	Total no. of ears	Observed CP rate (%)
Snow Jean	+2	22 Nov
Honey 236	+2	20 Nov	0.75	623	78	2.0746
			1.5	712	78	2.3710
			2.25	163	79	0.5359
			3	54	78	0.1798
			5.25	119	75	0.4121
			11.25	19	80	0.0617
			17.25	39	78	0.1299
			23.25	9	79	0.0296
Black Pearl	−3	25 Nov	0.75	226	78	0.7526
			1.5	261	78	0.8691
			2.25	38	79	0.1249
			3	10	78	0.0333
			5.25	6	75	0.0208
			11.25	1	80	0.0032
			17.25	0	78	0.0000
			23.25	0	79	0.0000
Tainung 1	−7	29 Nov	0.75	4	78	0.0133
			1.5	3	78	0.0100
			2.25	3	79	0.0099
			3	1	78	0.0033
			5.25	2	75	0.0069
			11.25	1	80	0.0032
			17.25	0	78	0.0000
			23.25	0	79	0.0000
Tainung 4	−10	2 Dec	0.75	1	78	0.0033
			1.5	4	78	0.0133
			2.25	0	79	0.0000
			3	0	78	0.0000
			5.25	0	75	0.0000
			11.25	0	80	0.0000
			17.25	0	78	0.0000
			23.25	0	79	0.0000

Table 6.

The mean of CP rate (%) by different temporal separation in in the first season

	Upwind				Downwind
Distance (m)	2 d early	Synchronous	5 d delayed	12 d delayed	2 d early	Synchronous	5 d delayed	12 d delayed
0.75	0.28	0.15	0.01	0.00	0.36	1.07	0.07	0.19
1.5	0.00	0.04	0.00	0.00	0.24	1.01	0.00	0.14
2.25	0.02	0.02	0.00	0.00	0.15	0.36	0.00	0.06
3	0.02	0.00	0.00	0.00	0.15	0.05	0.00	0.02
4.5	0.02	0.02	0.01	0.00
6	0.00	0.03	0.00	0.01
9	0.03	0.06	0.02	0.00
12	0.00	0.00	0.00	0.00

Table 7.

The mean of CP rate (%) by different temporal separation in the second season

	Upwind				Downwind
Distance (m)	2 d early	3 d delayed	7 d delayed	10 d delayed	2 d early	3 d delayed	7 d delayed	10 d delayed
0.75	0.19	0.34	0.00	0.00	2.07	0.75	0.01	0.00
1.5	0.31	0.10	0.01	0.00	2.37	0.87	0.01	0.01
2.25	0.28	0.09	0.00	0.00	0.54	0.12	0.01	0.00
3	0.14	0.10	0.01	0.01	0.18	0.03	0.00	0.00
4.5					0.44	0.25	0.01	0.00
6					0.18	0.00	0.00	0.00
9					0.23	0.02	0.01	0.00
12					0.12	0.00	0.00	0.00

Discussion

In Taiwan, the commercial maize varieties are hybrid cultivars. The CP between non-GM and non-GM maize does not lead to contamination problem of cultivars. On the contrary, the pollen-mediated gene flow from GM crops can result in the adventitious presence of GM material in conventional crop produces. The produces need to be labeled as being consisting of, containing or produced from a GM organism when the unavoidable presence of GM material in non-GM up to 5% in Taiwan and 0.9% in European Union (EU).

Furthermore, the target limit rate relates solely to GM regulations. However, due to the factor that no GM crops are currently planted in the open fields of Taiwan, this study provides an experimental reference in Taiwan and elsewhere with similar ecosystems.

Previous studies regarding the effect of synchrony during flowering time on the CP rate were employed by sowing the recipient or source plots on different dates. However, we used 4 local maize hybrid varieties with different flowering dates as pollen sources and one local maize hybrid variety as the recipient in this study. These source and recipient varieties were sown on the same day to discover the effect of non-coincidence flowering on the CP rate in maize at various distances on a field-scale level in Taiwan. This study shows that the synchronization of flowering time between the source plots and recipient field affects the extent of CP; thus, temporal separation may be a useful method for minimizing the CP rate between different maize hybrids. This result is consistent with those of other studies.^5,10,11,13

In this study, the average CP rates of the second season were higher than those of the first season. The reason may be that the seasonal conditions caused a different synchrony level in the 2 crop seasons. Moreover, the average wind speed was higher in the first season than in the second season, but the maximum wind speed was higher in the second season than in the first. Halsey et al.¹⁰ showed that wind peaks during pollen shedding rather than average wind speed are the force that affects pollen flow. Moreover, winds most commonly originated from the NW direction in the first season, but were actually from the N direction in the second season.

In this study, although Tainung No. 1 had a closer synchrony of flowering time with the pollen recipient than Tainung No. 4, Tainung No. 1 caused a lower CP rate than Tainung No. 4 downwind in the first season. The late flowering plants of the Snow Jean resulted in more exposure to the Tainung No. 4 than to Tainung No. 1 if the silk was unreceptive at the moment of fertilization because maize pollen remains viable only for 1 to 2 hr after shed.² Moreover, the instantaneous maximum wind speed during flowering had a more significant effect on gene flow.

Because the receptive silks of the recipient maize were saturated by the self pollen before the flowering of the source pollen, the pollen flow containment was more efficient in this study when the flowering dates of the pollen source varieties were later than those of the recipient variety. In the first season, the CP rate was <0.2% at a distance of 3 m when the flowering time for the recipient was 5 d later than that of the pollen source variety. In the second season, the CP rate was <0.02% at all distances when the flowering time for the recipient was 7 d later than that of the pollen source variety. The CP rate was <1% at a distance of 0.75 m when the flowering time was 3 d later. However, the varieties with closer synchrony resulted in a CP rate of >1% at a distance of 1.5 m and <1% at 2.25 m.

Because most agricultural fields in Taiwan are fragmented landscapes and spatially heterogeneous with small pollen sources and recipients, a coexistence strategy that only considers the isolation distance is impractical. Temporal separation and isolation distance can be considered together to minimize the adventitious presence of one crop with another. The results of this study can be a reference in establishing the coexisting systems of GM and non-GM crops.

Materials and Methods

Field design

Field experiments were performed in 2010 at the experimental field (24°1′ N, 120°41′ E) of Taiwan Agriculture Research Institute (TARI), Wufeng. The soil was a silt clay loam. The total area was approximately 0.45 ha (100 by 45 m) for the first season. Four equal-sized pollen source plots (20 by 15 m) separated from each other by a 4 m border were designated at the south direction (Fig. 3a). The total area was approximately 0.4575 ha (100 by 45.75 m) for the second season. Four identical source plots (20 by 15.75 m) separated from each other by a 4 m border were located at the north direction (Fig. 3b). The source plots were surrounded by the recipient field for the 2 crop seasons.

Plant materials and planting dates

Table 1 shows the agronomic characteristics of the plant materials. Four local varieties were used as pollen sources: Honey 236 with yellow wrinkled kernels, Black Pearl with purple round kernels, Tainung No. 1 with yellow round kernels, and Tainung No. 4 with white round kernels. The recipient maize was Snow Jean with white wrinkled kernels. We examined the kernel pericarp color and the shape of the maize, which results from the xenia effect. The xenia effect is found as a result of fertilization by different pollen sources after pollination. The effect of foreign pollen is immediately apparent in the endosperm-type maize type, including, size, color and shape.^19-21 Yellow wrinkled kernels appeared on the ears of the pollen recipient when it was pollinated by the Honey 236 pollen. Moreover, purple rounded kernels, yellow rounded kernels, and white rounded kernels appeared on the ears of the recipient when pollinated by the Black Pearl, Tainung No. 1, and Tainung No. 4 pollens, respectively (Fig. 4).

Figure 4.

The photograph of ears in 2010 experiment. (a) The pure maize ears of all 13 varieties. (b) Example of outcrossing from source plots (Black Pearl and 14 Honey 236) to receptor plots (Snow Jean). (c) Snow Jean arising in kernels 15 fertilized by Tainung No. 1. (d) Snow Jean arising in kernels fertilized by 16 Tainung No. 4.

Conventional farming practices were used in this experiment. The distance between the plants in a row was 0.3 m, whereas the distance between the rows was 0.75 m. The density was 44,400 plants ha⁻¹. The recipient and all pollen sources were planted on March 15 and September 29 for the first and second season of the experiments, respectively.

Climate monitoring during the flowing period

The meteorological information included wind speed and direction, and was recorded by a weather station that was located at the corner of the experiment field. Wind speed and direction were recorded from 6 a.m. to 4 p.m. with “hour” as the unit of record for 7 d before and after the silking time of the recipient (May 15, 2010 in the first season and Nov. 24, 2010 in the second season). Changes in wind speed and wind roses were used to interpret the trends in wind speed and direction.

Data collection and calculation of CP (%)

Each plant was tagged based on its orientation when the silk emerged from the sheath. Rows 1 to 60 were assigned from the first row on the south side to the last row on the north side during the first season (Fig. 3a). The ears of the recipient were sampled every 4 m from Rows 1 to 4. The pollen source plots were located between Rows 5 and 24; thus, the recipient plants in this region were sampled at distances of 0, 2, 4, 24, 26, 28, 48, 50, 52, 72, 74, 76, 96, 98, and 100 m from the west field edge to the east in Rows 7, 10, 13, 16, 19, and 22. Moreover, the ears of the recipient were sampled every 4 m from Rows 25 to 42 and rows with even numbers from Row 43 to the end of the field. The sampling schematic in the second season was similar to that of the first season, other than Rows 1 to 61, which were assigned from the first row on the north side to the last row on the south side (Fig. 3b). The ears of the recipient were sampled every 4 m from Rows 1 to 4. The pollen source plots were located between Rows 5 and 25; thus, the recipient plants in this region were sampled at distances of 0, 2, 4, 24, 26, 28, 48, 50, 52, 72, 74, 76, 96, 98, and 100 m from the west field edge to the east, in Rows 7, 10, 13, 16, 19, and 22. Moreover, the ears of the recipient were sampled every 4 m from Rows 26 to 42 and the rows with even numbers from Row 43 to the end of the field. In each sampled unit, the primary ears from 3 consecutive plants with the exception of missing plants were collected during both crop seasons.

Visual inspection of the pollen recipient’s ears was used to calculate the CP rate from different pollen sources. The kernel color and shape of the pollen recipient that was fertilized by the different pollen sources can be distinguished by phenotypic differences because of the xenia effect. Therefore, the CP rate (%) from the different pollen source can be calculated by counting the number of particular contaminated kernels on the ear of the recipient as follows:

$${\displaystyle \text{CP(!!\%)}=\frac{{\sum }_{i=1}^{n}Ea{r}_i}{n\times AVK}\times 100\%}$$

where n is the total number of ears at each sampling unit, Ear_i is the number of particular contaminated kernels on the i^th ear of the pollen recipient in each sampling unit, and AVK is the average total kernel number per ear in the whole field. One ear from each sampling unit was randomly selected to individually calculate the total number of kernels and obtain the AVK of the whole field. The overall CP rate was achieved by summing the CP rates from 4 pollen sources:

$${\displaystyle \text{overall CP(!!\%)}=\sum _{i=1}^{4}C{P}_i(\%)}$$

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.