Pollination timing effects on kernel set and silk receptivity in four maize hybrids.
Cell elongation requires maintenance of sufficient turgor ([[psi].sub.T]) to expand cells (Taiz and Zeiger, 1991). Hybrids have shown variability in the maintenance of silk [[psi].sub.T] during water deficits. Hybrid WF9 x A632 was able to maintain significantly higher silk [[psi].sub.T] when water stressed compared with B73 x MO17 (Schoper et al., 1987b). Anderson (1996) found that silk elongation decreased as silk water potential ([[psi].sub.w]) decreased. There were differing silk elongation patterns before and after recovery from water deficit for hybrids WF9 x A632 and B73 x MO17. Bassetti and Westgate (1993c) also reported relatively little silk growth at silk [[psi].sub.w] below -0.8 MPa for Pioneer 3790 and Pioneer 3732 in both field and greenhouse experiments. They found that water deficit can hasten silk senescence, causes ovary developmental failure, and that the timing of the water deficit is important in determining which of these most limits kernel number. Fertilization failure occurred because cells at the base of the silk collapsed (Bassetti and Westgate, 1993b). Data from Schoper et al. (1987b) indicate there is genetic variability in the ability to maintain silk [[psi].sub.T]. Adjustments in solute potential ([[psi].sub.s]) and cell wall elasticity are possible means of maintaining positive [[psi].sub.T] during water deficit (Volkenburgh and Boyer, 1985; Westgate and Boyer, 1985b). Inherent differences or regulation and adjustment of silk [[psi].sub.T] and silk water relations may impact silk growth, silk receptivity, and ultimately, kernel number and yield.
Differences in silk receptivity and [[psi].sub.T] may be due to genetic variability in the receptive area on a silk as well. The silk trichomes are the receptive area for pollen interception. They are especially penetrable by the pollen tube through the papillate tips because of a discontinuous cuticle (Kiesselbach, 1949; Heslop-Harrison et al., 1984). Rates of water loss from silks may be related to exposed trichome number or surface area. Cuticle discontinuities would also presumably allow for transpiration. Differences in silk receptive area might partially account for differences in silk receptivity, silk [[psi].sub.T], and silk health.
Stresses such as drought, high density, low fertility, and long photoperiods may impact silk growth rate and health. Selection for reduced anthesis-silk interval (ASI) as a secondary trait can improve tolerance to these stresses in a broad sense (Edmeades et al., 2000). Reduced ASI has corresponded to improvements in reproductive success across decades (Duvick, 1997). Reproductive success often corresponds to a negative correlation between ASI and grain yield under drought stress environments (Jensen, 1971). Silk growth dynamics, as it relates to ASI, may positively impact yield by optimizing kernel set. Successful hybrid kernel production may be related to the duration of silk receptivity. Understanding which hybrids can produce the most kernels when silks are exposed for a long period of time may be helpful in selecting for drought-stress tolerant hybrids and yield stability. Under drought conditions, there is often slow silk growth with final exposure several days after initial pollen shed. Long, functional silk life is desirable while silks are growing slowly under drought stress.
Seed set and kernel abortion are tied to maintaining a balance between sink demand and sink capacity for photosynthate. To maintain ovary growth, a steady flux of C is required (Zinselmeier et al., 2000). However, under stress situations, this balance is often upset.
The objective of our experiments was to explore the effects of timing of pollination on kernel number in four hybrids with different silk-water relationships or with suspected differences in yield stability. We considered yield stability relative to stresses such as water deficit that impact yield and flowering dynamics. It is hypothesized that genetic variability in silk growth rate and longevity of silk viability causes variability in kernel number when pollination timing is controlled relative to silk emergence. Additionally, these experiments quantify silk trichome morphology (surface area, number, length, and number per millimeter) to estimate the receptive area on the silks, and to evaluate variability for these traits among the four hybrids studied.
MATERIALS AND METHODS
Controlling Timing of Pollination
Four maize hybrids, B73 x MO17 (low silk [[psi].sub.T]), WF9 x A632 (high silk [[psi].sub.T]), Pioneer 3379 (stable dryland yields), and 3343 (unstable dryland yields), (Schoper et al., 1987b; Pioneer Hi-Bred International, 1989; Anderson, 1996) were hand-pollinated with excess pollen at 1, 2, 4, 6, 8, 10, 12, or 14 DAFS. The 6 DAFS treatment was omitted in 1993; the 12 and 14 DAFS treatments were only included in 1994. The randomized complete block (RCB) split-plot experimental design was replicated eight times in 1992 and four times in 1993 and 1994. Plots were in a maize-soybean [Glycine max (L.) Merr.] rotation in Johnston, IA, (Pioneer Hi-Bred International research farm) on a Wadena silt loam soil (fine-loamy over sandy, mixed, superactive, mesic Typic Hapludoll) with 3% organic matter. Planting dates were 11 May 1992, 14 May 1993, and 3 May 1994 for the first half of the experiments, with a final plant density of 69 000 plants [ha.sup.-1]. Replications 5 and 6 (1992) and 3 (1993 and 1994) were delayed 50 growing degree units (GDU) and planted by hand. Replications 7 and 8 (1992) and 4 (1993 and 1994) were delayed 100 GDU. Delayed replications were used to spread the workload at pollination time. Plots were fertilized with the equivalent of 157 and 134 kg [ha.sup.-1] N and K according to soil test recommendations.
The experimental unit was a 0.0024-ha six-row plot of maize on a 0.76-m row spacing. The subunit was the primary ear. Three ears were sampled for the kernel count. The experimental design included four hybrids (whole plot) and the treatments, randomized within each whole plot, were timing of pollination of primary ears (subplots) in each year. Treatments were applied (ears were pollinated) at various DAFS. Plots were hand-harvested at approximately 25% grain moisture (estimated by sampling open pollinated ears), and kernel number per primary ear (kernel set) was determined as the number of fertilized kernels filled and carried normally to physiological maturity. Kernel set did not include kernels that stopped grain fill during the linear phase of grain fill or earlier and equates to the final harvestable kernel number per ear. In 1993 and 1994, the number of aborted (shriveled) kernels, grain weight, and kernel weight were also quantified. A shriveled kernel refers to kernels that had been fertilized but did not develop normally to physiological maturity. Salvador and Pearce (1995) define hypoplasty as subnormal growth or growth cessation of sexually fertilized kernels. These kernels accumulated some starch, exhibited some early growth, and had a shriveled appearance at physiological maturity. Aplastic kernels that never accumulate starch were not evaluated or kept separate.
Ear shoots were covered before silking with bags to prevent random pollination. Bags were marked with the date when silks were first visible (0 DAFS). Since plots were monitored daily, all of the exposed silk growth at 0 DAFS occurred some time during the previous 24 h. Hand pollinations were made on predetermined DAFS for each treatment with excess pollen from neighboring plants or from delay-seeded plants growing nearby. The pollinating bags remained over ear shoots after pollination so that pollen was only available at the time of treatment. In addition, three control plants were marked and left with silks exposed to allow open pollination.
Silk Elongation Measurements
Silk elongation rate was taken on exposed silks in 1992. Silks from two plants per plot were bundled into three groups of eight to 12 silks per ear in the eight replications. A total of 48 silk-length observations were collected at each timing treatment for each hybrid across replications. Silks were marked with small strips of tape near the apex and measured on the day of first appearance (DAFS = 0). Measurements were made by pulling the bundled silks tight and using a ruler to measure the distance from silk apex to point of silk emergence at the shoot husk. We tried to minimize tissue handling and determine silk length consistently. It is possible that repetitive handling and bundling the silks affected growth somewhat. The cob rachis and husk also grow each day, potentially causing variability in the reference point for measuring silks. The same silk bundles were measured each day. Consequently, measurements represented the earliest emerging silks, that is, those from the midbase region of the ear, and do not include the dynamics of successive emergence of silks from different floret positions. The growth rate at 0 DAFS is only an estimate because we do not know at what time in the previous 24 h the silks emerged from the husk. Measurements were made each day for 11 d at 0700 h and were completed by 0830 h to ensure they were not pollinated during the measuring process. Pollen dehiscence generally occurs later in the morning as temperatures increase and morning dew evaporates. Silks were kept covered with bags between measurements to prevent pollination.
Subsamples of silks were collected at 3 DAFS in 1993 and 1994. Within each of the four replications, two silks which were randomly selected from a bulk of silks from four ears were examined for silk trichome number per centimeter, length, and total number. To examine potential differences across the length of the silk, the exposed portion of the silk was evenly divided into thirds, and two measurements were taken on each segment. The apex was the tip end segment of the exposed silk, mid was the middle third of the exposed silk, and the base segment was the exposed portion nearest the ear shoot (silk exit point from shoot husk).
Measurements were made under an optical microscope (Olympus Optical Co. Ltd., Japan) at 100x magnification. A slide micrometer, in combination with an ocular lens micrometer, was used to calibrate the actual dimensions in the field of view. Only straight, extended, trichomes were measured for length. Receptive area was estimated by assuming silk trichomes were conical in shape.
The ANOVAs indicated nonsignificant effects due to planting date, so planting dates were treated as replications (blocks) in each of the 3 yr. An ANOVA using the RCB split-plot design was performed for each year using pooled means from three plants. The block x treatment error was used to estimate Error A and the pooled error was used to estimate Error B for treatments and interactions. The combined years ANOVA used the pooled means for hybrid x environment due to unbalanced replication (P. Hinz, 1994, personal communication). Only treatments at 1, 2, 4, 6, 8, and 10 DAFS were used in the combined analysis and the 6 DAFS treatment was estimated in 1993 using the General Linear Models procedure (SAS Institute, 1985). The combined years ANOVA used environment x hybrid as Error A and the pooled error to estimate treatments and interactions.
RESULTS AND DISCUSSION
Silk Growth Rate
Silk elongation rates were evaluated across hybrids. Hybrids measured for silk elongation of the ear's midbase position showed a general pattern of high initial growth rate at 1 DAFS (averaging 41.6 mm [d.sup.-1]) and 2 DAFS (45.0 mm [d.sup.-1]), followed by a gradual decrease and leveling in growth rate at 3 to 10 DAFS (Fig. 1). By 4 DAFS, growth rate had dropped 43%; by 6 DAFS, 69%; by 8 DAFS, 89%; and by 10 DAFS, the rate of growth was negligible. The total mean silk length across hybrids was 233 mm. (Table 1).
There were significant differences among hybrids for silk length beyond the husk tip at 10 DAFS. Shorter and slower elongation of B73 x MO17 silks relative to the other three hybrids contributed to the significant time x hybrid interaction. Hybrid B73 x MO17 had the shortest silks at 207 mm followed by WF9 x A632 at 228 mm. Although 3379 attained the longest silks (252 mm at 10 DAFS), it did not have as high an initial rate of growth (0-3 DAFS) as 3343. Differences in husk length between the hybrids could affect silk length measurements. However, hybrid husk length differences appeared to be minimal relative to silk elongation based on general observations made while measuring the silks.
Unlike the other three hybrids, WF9 x A632 exhibited a somewhat different growth pattern with its fastest growth occurring on the day silks first emerged (Fig. 1). The longer initial silk lengths for WF9 x A632 resulted in a larger percentage of the ear's silks and surface area exposed to the desiccating environment outside the husk for a longer period of time in this hybrid. WF9 x A632 silk growth rate declined from the day of silk emergence. Although WF9 x A632 had the fastest initial silk growth rate, it has been shown to maintain high silk [[psi].sub.T] under water deficit stress, as well as better silk growth recovery after relief from water deficit (Schoper et al., 1987b; Anderson, 1996) relative to B73 x MO17. Fast silking may offer some advantages by increasing silk exposure to shedding pollen, and this could lead to increased kernel set under conditions of protandry (pollen shed before silk emergence), as can occur in drought environments when the interval between pollen shed and silk emergence is increased (Johnson and Herrero, 1981; Hall et al., 1982; Bassetti and Westgate, 1994). Fast silking may provide a more effective reproductive strategy than slow silking hybrids by exposing a larger percentage of potential silk area earlier than other hybrids. However, we do not offer evidence that WF9 x A632 grain yield is more tolerant to drought or more stable than B73 x MO17 in normal commercial field environments.
[FIGURE 1 OMITTED]
Kernel Set Dynamics
A significant time x hybrid interaction indicated that WF9 x A632 lost silk receptivity at a significantly faster rate than the other three hybrids (linear slope = -0.9 kernels [d.sup.-l]). As expected, the relationship between kernel number and pollinating time was nonlinear [F (4, 36) = 14.4, P = 0.001]. Even so, the linear component of the variance MS was large in all analyses run, and slopes from the linear analysis should be adequate for a general separation of these large interactions (time x hybrid) (P. Hinz, 1994, personal communication). Hybrid WF9 x A632 clearly had the greatest kernel number when pollinated early and dropped off (due to silk senescence) in all 3 yr more rapidly than the other hybrids (15.7 kernels [d.sup.-1] in 1992, -27.6 in 1993, and -16.4 in 1994). All 3 yr showed significant time x hybrid interactions and hence differences between slopes (Fig. 2).
[FIGURE 2 OMITTED]
The maximum kernel number for WF9 x A632 was only maintained for a few days, indicating less stability for kernel number when pollen is unlimited compared with the other hybrids (Fig. 2). As with all hybrids, stability would decrease even more if pollen supply were limiting (Flottum et al., 1984; Bassetti and Westgate, 1994). Successful pollination of maximum number of florets and maintenance of maximum kernel number for a long duration is critically important in determining kernel set in environments such as drought where stress may delay silk emergence. (Sadras et al., 1985a, 1985b).
Hybrids 3379, 3343, and B73 x MO17 had greater kernel set when pollinated 10 DAFS or later in 1992, 1993, and 1994, compared with WF9 x A632 (Fig. 2). These three hybrids maintained maximum kernel number with later pollination, resulting in greater stability for kernels per ear. Because hybrids responded differently with years, we cannot predict the DAFS when receptivity will begin to decline for any given hybrid in any given year. However, we can note the time when maximum kernel number was observed in each year for these experiments (Fig. 2). It may be valuable to characterize hybrid variability in silk receptivity as a trait to consider for kernel set stability.
Understanding genotype x environment interactions may help with optimization of pollination synchrony. Although the pattern for hybrid kernel set and silk senescence was broadly similar in 1992, 1993, and 1994 across pollinations, the rate of silk emergence, the duration of maximum kernel set, and the rate of decline in silk receptivity varied among hybrids by year. Furthermore, the DAFS when kernel set first declined was different in some years. For example, in 1992 and 1993, WF9 x A632 had the highest kernel set when pollinated at 1 DAFS compared with the other three hybrids, presumably because it produces most of its silk in one large group. In 1992, WF9 x A632 kernel set declined when ears were pollinated after 8 DAFS while 3343, 3379, and B73 x MO17 maintained maximum kernel set at 10 DAFS (Fig. 2A). In 1993, kernel set declined for pollinations between 4 and 8 DAFS for WF9 x A632, after 8 DAFS for 3343, while B73 x MO17 and 3379 again maintained maximum kernel set at 10 DAFS (Fig. 2B). In 1994, time of pollination was extended out to 14 DAFS, revealing a significant decline in kernel set when pollination was delayed for all four hybrids (Fig. 2C). However, although silk elongation was fast at emergence, WF9 x A632 lost silk receptivity earlier and faster than the other three hybrids. The fact that silk receptivity lasts so long in general is key, and indicates that under most circumstances silk dynamics may not limit potential. Managing or selecting for the kernels per ear increase associated with delayed pollinations may justify the risk of losing some silk viability. Protogynous silking (up to 3 to 4 d before pollen shed) would appear to be a good approach to increasing kernels per ear under optimum growing conditions. However, protogynous silking in some hybrids results in significant silk tissue growth that may potentially shield older silks from pollen interception. The vigorous hand pollinations with excess pollen could have masked this potential physical barrier to pollen interception.
By looking at kernel set across years we can make some interesting observations about optimizing kernel set. Hybrids averaged 400 kernels per ear when pollinated at 1 DAFS and kernel set increased to a maximum of 621 by 8 DAFS. This interval (1 to 8 DAFS) represents the time when exposed silk area was limiting to kernel set. However, the ranking of the means for each hybrid changed at different times of pollination. This corresponds to a different silk elongation pattern and rate of loss in silk function as silks age. Hybrid means indicate WF9 x A632 silked quickly and had the greatest kernel number (454) when pollinated at 1 DAFS, followed by B73 x MO17 (401), 3379 (387), and then 3343 (356). Hybrid WF9 x A632 kernel number fell rapidly as pollination was delayed. Hybrid 3343 reached maximum kernel number (640) when pollinated on 8 DAFS. Hybrid 3343 had 14 to 13% fewer kernels per ear (356-521) at 1 to 4 DAFS than the other three hybrids (414-599). The mean maximum kernel number across all four hybrids was 645 kernels per ear in 1992 and occurred at 10 DAFS, whereas in 1993 it was 587 on 4 DAFS. In 1994, the maximum occurred at 8 DAFS (654).
Interestingly, for these hybrids, open pollinated controls often resulted in fewer kernels than the maximum achieved with one-time pollination (Fig. 2). In 1994, the 3343 open-pollinated primary ears achieved 85% of the maximum kernels [ear.sup.-1] that was obtained on 8 DAFS with synchronous pollination. Hybrid 3379 obtained 88% of the maximum for synchronous pollination on 8 DAFS, B73 x MO17 obtained 91% of the maximum at 8 DAFS, and WF9 x A632 obtained 82% of the maximum on 6 DAFS. Similar hybrid rankings occurred in 1992 and 1993, but in some cases the maximum occurred at differing DAFS. Carcova et al. (2000) also found that synchronous pollination was advantageous. Pollination at 5 DAFS improved the kernel number per plant over natural pollination on apical ears for hybrids 'Dekalb DK 752', 'DK 664', 'AgriPro AP 162', and 'AP 9191'. Furthermore, they found that the floret fertility index (FFI = number of kernels/number of pollinated silks) improved, and the extent of kernel set and FFI improvement was hybrid dependent. Delayed fertilization of early silking ovaries apparently allowed later-developing flowers to achieve greater potential, resulting in more successful tip kernel pollination. In our experiments, by 8 to 10 DAFS all silks were exposed and subsequent decreases in kernel number were likely due to reduced silk receptivity. Seed set and kernel abortion are tied to maintaining a balance between sink demand for a continuous stream of photosynthate and sink capacity for photosynthate relative to kernel initiation, growth, and development requirements associated with synchronous or asynchronous pollination (Zinselmeier et al., 2000). Continuous pollination across time may upset the steady balance between sink demand and sink capacity more than synchronous pollination. It may be of commercial value to consider the possibility that kernels per ear may be optimized by delayed pollination.
Although our experiments were not done with inbreds, poor synchronization of pollen shed with silking in seed fields could result in situations where pollen becomes available when silks are older and less receptive, in a situation analogous to protogyny. Maximum kernel set will require pollen availability at sufficient intensities as well as optimum pollination timing (Bassetti and Westgate, 1993b, Carcova et al., 2000). There is limited information available on what the threshold is for optimum pollen shed intensity. Bassetti and Westgate (1994) determined with computer-aided imaging analysis that in hybrid ears with silks exposed to pollen for 1 d or more at intensities >100 grains [cm.sup.-2] [d.sup.-l], nearly all florets set kernels.
Yield, Kernel Abortion, and Silk Age
In 1993 and 1994, the number of kernels that were aborted and shriveled were visually determined by differentiating poor embryo and endosperm development from unpollinated floret positions (kernel reduction due to unexposed silks, loss in silk receptivity, or very early arrest of kernel growth with no starch accumulation). The maximum number of shriveled kernels occurred on ears pollinated at 4 DAFS on average. There was differential shriveling between hybrids (Fig. 3). Kernel reduction due to shriveling as defined above amounted to a mean loss of 11.9 kernels per ear, or a 2% loss. The highest average kernel shriveling was for WF9 x A632 (17.9 kernels) and for 3343 (17.1), followed by 3379 (6.4) and B73 x MO17 (5.5).
[FIGURE 3 OMITTED]
By examining the integrity of the endosperm and embryo, we used kernel shriveling in our attempt to identify kernels that had been fertilized but did not develop normally to physiological maturity. Subnormal growth in which some starch may accumulate was termed hypoplasty by Salvador and Pearce (1995). Shriveled kernels are fertilized and represent a component of kernel abortion separate from loss of silk function where fertilization is unsuccessful. Another component of arrested kernel growth where fertilization may be unsuccessful is described as being aplastic. Starch does not accumulate in aplastic kernels. However, this component was not differentiated in the shriveled kernel category (Salvador and Pearce, 1995). Some kernel shriveling occurred at each time of pollination. Bassetti and Westgate (1993c) suggested that kernel abortion at 2 to 3 d after pollination may be the cause for reduced kernel set under low [[psi].sub.w]. The method used to evaluate kernel shriveling in our experiments may fail to detect such early embryonic developmental failure and differentiate it from loss of silk receptivity. We measured loss of kernels after the linear phase of filling began and the physical outcome of these kernels may have been determined in the dilatory or exponential phase of grain fill. The dilatory phase of kernel development is a period of slow dry matter accumulation soon after fertilization. The exponential phase occurs when normal kernels initiate rapid development and aborted hypoplastic kernels are left behind (Salvador and Pearce, 1995). Kernel loss due to kernel shriveling in our experiments appeared predominately at the apical position of the ears, whereas reduction in kernels due to loss in silk receptivity was progressive from ear base to apex (Peterson, 1949; Bassetti and Westgate, 1993a, 1993b). Decreased silk receptivity was a larger component contributing to kernel loss than kernel shriveling as defined in our experiments.
Carcova and Otegui (2001) found that pollination gaps of 2 and 4 d after 2 d of open pollination reduced kernel set for hybrids DK752 and DK664, whereas a gap of 6 d resulted in better kernel set. This indicates some type of interference of early pollinated silks on late pollinated flowers apparently unrelated to reduced silk receptivity, and occurring when there is a lack of synchrony in pollination. In our hand-pollinated treatments, only single pollinations were used so the demands for assimilates were not influenced by a previous pollination as would occur in the open-pollinated control treatment. To maintain ovary growth, a steady flux of C is required (Zinselmeier et al., 2000). Furthermore, with use of single controlled-timing pollination, we observed that midbase position silks were exposed longest for WF9 x A632 and kernels at these positions were the first to discontinue successful kernel development, perhaps indicating that length of silk exposure to desiccating environment may be a primary factor determining duration of silk receptivity. Drought may exacerbate this process since Bassetti and Westgate (1993c) found that water deficit stress hastened silk senescence, resulting in limited kernel set. In nonstress conditions, flowers from the midbase region of the ear failed to set kernels seven or more days after their silks first emerged. Fertilization failure occurred because cells at the base of the silk collapsed (Bassetti and Westgate, 1993b).
The weight of the grain (yield) from the primary ear was positively correlated (r = 0.73) with kernel number and negatively correlated with the weight per kernel (r = -0.34). Field weight and kernel number reached maximum in ears pollinated between 4 and 6 DAFS (Fig. 4). Contrastingly, the weight per kernel was at a maximum for earliest pollination and declined to a near-constant weight for ears pollinated by 6 to 10 DAFS. The majority of the change in grain weight was likely attributable to kernel number because silk number was limiting the number of floret positions that could be pollinated early, and loss of silk receptivity causes reduced kernel set for late pollinations (Bassetti and Westgate, 1993a, 1993b). The weight of individual kernels was greater for early pollination perhaps because these kernels were from the midbase ear position that normally produces larger kernels. Early pollinations also produced lower kernel number, which may permit increased carbohydrate partitioning to those kernels and extend the filling period by a few days. Otegui and Melon (1997) suggest that assimilate partitioning to the ear during silking is critical to kernel set, and is related to the rate of ear growth (in length) around silking.
[FIGURE 4 OMITTED]
Silk Receptive Area
We found that hybrids differ in silk trichome length, number per millimeter of silk length, and surface area, similar to how trichomes on leaves may differ among cultivars (Sayre, 1920; Traore et al., 1989). The number of silk trichomes per millimeter length of WF9 x A632 was less than 3343 and 3379 in 1993 and less than the other three hybrids in 1994 (Fig. 5). Additionally, total trichome number on exposed silk surfaces was also lower for WF9 x A632, resulting in a much lower estimate of receptive area. A t test for linear contrasts of WF9 x A632 mean trichome surface area vs. the mean for the other three hybrids as a group was significant both years. The surface area differences of the silk trichomes may contribute to genetic variation for failure to set kernels with silk age (Heslop-Harrison et al., 1984).
[FIGURE 5 OMITTED]
The hybrid B73 x MO17 had the shortest exposed silk length and the longest trichome length of the four hybrids. A close examination of the trichomes of B73 x MO17 and WF9 x A632 revealed that WF9 x A632 trichomes were shorter and appeared to have fewer cells than B73 x MO17. The apparent thinning or discontinuity of the cuticle at the ends of the trichomes in both hybrids was also visible (Fig. 6). The number and lengths of trichomes may physically affect the efficiency of silks in capturing pollen and promoting successful pollen tube growth, especially when pollen becomes limiting (Kiesselbach, 1949; Heslop-Harrison et al., 1984). The unique differences in the silk trichome characteristics of WF9 x A632 may be related to its early loss of silk receptivity relative to the other hybrids.
[FIGURE 6 OMITTED]
Silk and trichome surface area may affect silk transpiration rates, allowing for cooling of silk and ear-shoot cells and maintenance of silk [[psi].sub.T]. However, the extra capacity of silks with many trichomes or highly receptive surface area to capture pollen may well offset any possible effects of loss of [[psi].sub.T]. The silk trichomes are covered by a discontinuous cuticle (Heslop-Harrison et al., 1984). The extent of the discontinuity may influence silk transpiration similar to how aperture size and stoma number influence transpiration in leaves (Williams, 1950; Bange, 1953; Jarvis and Mansfield, 1981). The thickness and the hydrophobicity of the wax layer comprising the silk cuticle is unique (Yang et al., 1994), and may also influence the duration of silk receptivity. All of these factors may contribute to regulating differences in rate of silk elongation, maintenance of silk [[psi].sub.T], as well as how long silks remain receptive, and they are the focus of other experiments (Anderson, 1996).
We also examined trichome length in various regions of the exposed silks at 3 DAFS. Trichomes were longest near the apex of the silk and were shorter as we moved toward the base of the exposed silk (Fig. 7). The silk apex cells are the most mature and likely have experienced more complete cell expansion and cell wall extensibility, resulting in longer trichome length at 3 DAFS. Trichomes were longest for B73 x MO17. Trichome length was greater in 1994 than in 1993, but all hybrids tended to follow the same trend of longer trichomes at the silk apex. Differences in trichome length along the position of the exposed silk may influence the receptive area and success of intercepting pollen under limiting pollen availability. The hybrid and spatial variation in lengths of trichomes along the exposed silks could potentially influence silk transpiration and silk [[psi].sub.T] by influencing the rate of water loss differentially along the length of the silk.
[FIGURE 7 OMITTED]
There is hybrid variability in silk morphology traits that may impact reproductive success and stability. Perhaps the high receptive surface area for B73 x MO17 is advantageous, resulting in pollination stability. Certainly, the exposed surface area of trichomes could impact the likelihood of physically intercepting a pollen grain in certain environments. More research is needed to address these interesting questions.
Time of pollination x hybrid interactions for kernels per ear were primarily a function of different hybrid rates of silk exposure and loss of silk receptivity with aging. The hybrid with the fastest initial silk growth rate (WF9 x A632) consistently lost silk receptivity sooner, and kernels per ear declined faster compared with B73 x MO17, 3379, and 3343. Conditions of protandry or protogyny will favor maximum kernel number in some hybrids more than others because of different silk emergence-senescence patterns, and different times when pollen intensity and shed duration become limiting to grain yield. Hybrid WF9 x A632 reached maximum kernel number when pollination occurred shortly after first silk emergence, and this resulted in a hastened loss of silk receptivity. Hybrids B73 x MO17, 3379, and 3343 reached maximum kernel number when pollinated 2 to 4 d later than WF9 x A632. Therefore, any delay in silk emergence caused by environmental stress that slows silk growth is likely to shift these response curves downward, since receptivity is likely a function of silk age. Hybrid WF9 x A632 had a corresponding lower trichome surface area, number, and number per millimeter than B73 x MO17, 3379, or 3343. The silk trichome length varied somewhat along the exposed length of the silk, and this may impact the efficiency of successful pollen interception and hence kernel set. The duration of time that maximum kernel number can be maintained under unlimited pollen supply may be an important trait to consider in selecting for yield stability in maize hybrids and in management of production seed fields.
Table 1. Silk length means for hybrids during 10 days after first silk (DAFS) in 1992. The LSD (0.05) for differences between treatment means for a given hybrid = 10.5 mm. The LSD (0.05) for differences between hybrid means on a given day is 15.4. Time of pollinar (DAFS) Hybrid 0 1 2 mm 3343 37 90 148 3379 33 78 127 B73 x M017 27 62 95 WF9 x A632 46 79 119 Mean 35 77 122 Rate, mm [d.sup.-1] 35.0 41.6 45.0 Hybrid 3 4 5 6 mm 3343 190 213 232 241 3379 167 192 212 230 B73 x M017 124 151 164 182 WF9 x A632 146 171 187 196 Mean 157 182 199 212 Rate, mm [d.sup.-1] 34.9 24.7 17.1 13.3 Hybrid 7 8 9 10 mm 3343 252 254 250 245 3379 241 248 250 252 B73 x M017 195 202 204 207 WF9 x A632 207 211 222 228 Mean 224 229 231 233 Rate, mm [d.sup.-1] 11.9 4.8 2.4 1.6
The authors thank Greg Edmeades, Antonio Perdomo, Jeff Schussler, and Chris Zinselmeier for their assistance in critically reviewing this research. We thank Dr. Harry T. Horner (Director of Bessey Microscopy Facility, Iowa State Univ., Ames, IA) for assistance in optimizing images. We would like to acknowledge Robert Whitmoyer (Ohio Agricultural Research and Development Center) for his assistance and expertise in image collection. We gratefully acknowledge the support of Pioneer Hi-Bred Int. Inc. (Johnston, IA).
Abbreviations: [[psi].sub.T] turgor; [[psi].sub.w], water potential; ASI, anthesis-silk interval; DAFS, days after first silk; FFI, floret fertility index: GDU, growing degree units; RCB, randomized complete block.
Anderson, S.R. 1996. Silk growth, cuticular wax composition and silk receptivity in four maize hybrids. M.S. thesis. Iowa State Univ., Ames (ISU 1996 A554).
Bange, G.J. 1953. On the quantitative explanation of stomatal transpiration. Acta Bot. Need. 2:255-296.
Bassetti, P., and M.E. Westgate. 1993a. Emergence, elongation, and senescence of maize silks. Crop Sci. 33:271-275.
Bassetti, P., and M.E. Westgate. 1993b. Senescence and receptivity of maize silks. Crop Sci. 33:275-278.
Bassetti, P., and M.E. Westgate. 1993e. Water deficit affects receptivity of maize silks. Crop Sci. 33:279-282.
Bassetti, P., and M.E. Westgate. 1994. Floral asynchrony and kernel set in maize quantified by image analysis. Agron. J. 86:699-703.
Carcova, J., and M.E. Otegui. 2001. Ear temperature and pollination timing effects on maize kernel set. Crop Sci. 41:1809-1815.
Carcova, J., M. Uribelarrea, L. Borras, M. E. Otegui., M.E. Westgate. 2000. Synchronous pollination within and between ears improves kernels set in maize. Crop Sci. 40:1056-1061.
Duvick, D.N. 1997. What is yield? p. 332-335. In G.O. Edmeades et al. (ed.) Developing drought and low-N tolerant maize. CIMMYT, El Batan, Mexico.
Edmeades, G.O., J. Bolanos, A. Elings, J.-M. Ribaut, M. Banziger, and M.E. Westgate. 2000. The role and regulation of the anthesis-silking interval in maize, p. 43-73. In M.E. Westgate and K.J. Boote (ed.) Physiology and modeling kernel set in maize. CSSA Spec. Publ. No. 29. CSSA and ASA, Madison, WI.
Flottum, P.K., D.C. Robacker, and J.E.H. Erickson. 1984. A quantitative sampling method for airborne sweet corn pollen under field conditions. Crop Sci. 24:375-377.
Hall, A.J., F. Vilella, N. Trapani, and C. Chimenti. 1982. The effects of water stress and genotype on the dynamics of pollen-shedding and silking in maize. Field Crops Res. 5:349-363.
Heslop-Harrison, Y., B.J. Reger, and J. Heslop-Harrison. 1984. The pollen-stigma interaction in the grasses. Tissue organization and cytochemistry of stigma (silk) of Zea Mays L. Acta Bot. Neerl. 33:81-99.
Jarvis, P.G., and T.A. Mansfield. 1981. Stomatal physiology. Cambridge Univ. Press, Cambridge.
Jensen, S.D. 1971. Breeding for drought and heat tolerance in corn. p. 198-208. In J.I. Sutherland and R.J. Falasca. (ed.) Proc. 26th Annu. Corn and Sorghum Ind. Res. Conf., Chicago. 14-16 Dec. 1971. Am. Seed Trade Assoc., Washington DC.
Johnson, R.R., and M.P. Herrero. (1981). Corn pollination under moisture and high temperature stress, p. 66-77. In H.D. Loden and D. Wilkinson (ed.) Proc. 36th Annual Corn and Sorghum Industry Res. Conf., Chicago. 9-11 Dec. 1981. American Seed Trade Assoc., Washington, DC.
Kiesselbach, T.A. 1949. The structure and reproduction of corn. Bull. 161. Univ. Nebraska Agric. Exp. Res. Stn., Lincoln, NE.
Mitchell, J.C., and J.F. Petolino. 1988. Heat stress effects on isolated reproductive organs of maize. Plant Physiol. 133:625-628.
Otegui, M.E., and S. Melon. 1997. Kernel set and flower synchrony within the ear of maize: I. Sowing date effects. Crop Sci. 37:441-447.
Peterson, D.P. 1949. Duration of receptiveness in corn silks. Agron. J. 34:369-371.
Pioneer Hi-Bred International. 1989. 1989 Pioneer brand products Northland sales area catalog. Pioneer Hi-Bred Int., Johnston, IA.
Sadras, V.O., A.J. Hall, and T.M. Schlichter. 1985a. Kernel set of the uppermost ear in maize: I. Quantification of some aspects of floral biology. Maydica 30:3747.
Sadras, V.O., A.J. Hall, and T.M. Schlichter. 1985b. Kernel set of the uppermost ear in maize: II. A simulation model of effects of water stress. Maydica 30:49-66.
Salvador, R.J., and R.B. Pearce. 1995. Proposed standard system of nomenclature for maize grain filling events. Maydica 40:141-146.
SAS Institute. 1985. SAS user's guide: Statistics. 5th ed. SAS Inst., Cary, NC.
Sayre, J.D. 1920. The relation of hairy leaf covering to the resistance of leaves to transpiration. Ohio J. Sci. 20:55-86.
Schoper, J.B., R.J. Lambert, and B.L. Vasilas. 1986. Maize pollen viability and ear receptivity under water and high temperature stress. Crop Sci. 26:1029-1033.
Schoper, J.B., R.J. Lambert, and B.L. Vasilas, 1987a. Pollen viability, pollen shedding, and combining ability for tassel heat tolerance in maize. Crop Sci. 27:27-31.
Schoper, J.B., R.J. Lambert, B.L. Vasilas, and M.E. Westgate. 1987b. Plant factors controlling seed set in maize. Plant Physiol. 83:121-125.
Taiz, L., and E. Zeiger. 1991. Plant physiology. Benjamin/Cummings Publ. Co., Redwood City, CA.
Traore, M., C.Y. Sullivan, R.J. Rosowski, and K.W. Lee. 1989. Comparative leaf surface morphology and the glossy characteristic of sorghum, maize and pearl millet. Ann. Bot. (London) 64:447-453.
Undersander, D.J. 1987. Yield and yield component response of maize to water stress in hybrids with different sources of stress tolerance. Maydica 32:49-60.
Volkenburgh, E.V., and J.S. Boyer. 1985. Inhibitory effects of water deficit on maize leaf elongation. Plant Physiol. 77:190-194.
Westgate, M.E., and J.S. Boyer. 1985a. Carbohydrate reserves and reproductive development at low leaf water potentials in maize. Crop Sci. 25:762-769.
Westgate, M.E., and J.S. Boyer. 1985b. Osmotic adjustment and the inhibition of leaf, root, stem and silk growth at low water potentials in maize. Planta 164:540-549.
Westgate, M.E., and J.S. Boyer. 1986. Reproduction at low silk and pollen water potentials in maize. Crop Sci. 26:951-956.
Williams, W.T. 1950. Studies in stomatal behavior: IV. The water relations of the epidermis. J. Exp. Bot. 1:114-131.
Yang, G., B.R. Wiseman, and K.E. Espelie. 1994. Effect of cuticular lipids from silks of selected corn genotypes on the development of corn earworm larvae. J. Entomol. Sci. 29:239-246.
Zinselmeier, C., J.E. Habben, M.E. Westgate, and J.S. Boyer. 2000. Carbohydrate metabolism in setting and aborting maize ovaries. p. 1-14. In M.E. Westgate and K.J. Boote (ed.) Physiology and modeling kernel set in maize. CSSA Spec. Publ. No. 29. CSSA and ASA, Madison, WI.
Steven R. Anderson, * Michael J. Lauer, John B. Schoper, and Richard M. Shibles
S.R. Anderson, 7301 NW 62nd Ave., M.J. Lauer, 6900 NW 62nd Ave., and J.B. Schoper, 7300 NW 62nd Ave., Pioneer Hi-Bred International, Johnston, IA 50131-1004; R.M. Shibles, Dep. of Agronomy, Iowa State Univ., Ames, IA 50011-1010. Research supported by Pioneer Hi-Bred International, Inc., Johnston. IA. Part of a thesis submitted by S.R. Anderson in partial fulfillment of requirements of a M.S. degree at Iowa State University. Received 11 Feb. 2003. * Corresponding author (firstname.lastname@example.org).
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Crop Physiology & Metabolism|
|Author:||Anderson, Steven R.; Lauer, Michael J.; Schoper, John B.; Shibles, Richard M.|
|Date:||Mar 1, 2004|
|Previous Article:||Legal liability issues in agricultural biotechnology.|
|Next Article:||Climatic and water availability effects on water-use efficiency in wheat.|