EFFECTS OF SOIL NITROGEN ON POLLEN PRODUCTION,POLLEN GRAIN SIZE,AND POLLEN PERFORMANCE IN CUCURBITA PEPO (CUCURBITACEAE)

To determine the effects of soil nitrogen on pollen production, pollen size, and pollen performance, two cultivars of zucchini (Cucurbita pepo) were grown under two nitrogen regimes in an experimental garden. The two cultivars were true breeding for alternative alleles for a one gene trait, ovary color. The soil nitrogen treatment had a significant effect on most measures of reproductive output through the female function. The nitrogen treatment did not affect the number of staminate flowers per plant but did have an effect on the number of pollen grains per staminate flower and the mean pollen grain size. A pollen mixture experiment revealed that pollen produced by plants in the high nitrogen treatment sired significantly more seeds than pollen from low nitrogen plants. Moreover, we found that the high nitrogen pollen sired even a greater percentage of seeds in the region of the fruit (ovary) previously shown to be fertilized by the fastest growing pollen tubes. Thus, the difference in the number of seeds sired by pollen from the two nitrogen treatments is due to differences in pollen performance. We conclude that spatial heterogeneity in soil nitrogen can influence the paternity of seeds in a plant population.     

Order of fertilization within the ovary in Phaseolus coccineus L. (Leguminosae)

Summary Phaseolus coccineus typically has six linearly arranged ovules per ovary. The three ovules near the stylar end of the fruit (positions one, two, and three) are more likely to produce mature seeds, to produce heavier seeds, and to produce more vigorous progeny than the ovules in positions near the peduncular/basal end of the fruit (ovule positions four, five, and six). We conducted a series of field experiments designed to supplement our understanding of the mechanisms determining these position effects. We found that approximately 98% of the ovules in 752 fruits were fertilized — about 0.6% of the stylar ovules were not fertilized, whereas 3.2% of the basal ovules were unfertilized. Moreover, we found that only about 49% of the ovules in these 752 fruits produced mature seeds. Over 60% of the stylar ovules produced mature seeds, whereas only 37% of the basal ovules produced mature seeds. Consequently, the proportion of fertilized ovules cannot explain the differences in seed maturation among the ovule positions. We found that after 6.5 h most of the fertilized ovules were located in the stylar ovule positions, and that there were no fertilized ovules in ovule positions five and six, indicating that the stylar ovules are fertilized first. When only the fastest growing pollen tubes were permitted to enter the ovary (due to exision of the style), only the ovules at the stylar end were fertilized, indicating that the ovule positions that are fertilized first are indeed fertilized by the fastest growing pollen tubes.On leave from the Escuela de Biologia, Universidad de Costa Rica, Cuidad Universitaria Rodrigo Facio, San Jose, Costa Rica, Central America     

Microgametophytic selection in two alfalfa (Medicago sativa L.) clones

Summary Microgametophytic selection was investigated using two ecologically diverse autotetraploid clones of alfalfa. Several selection pressures (drying, aging, freezing, and high and low temperatures) were applied to microgametophytes at three stages of the life cycle, 1) during microsporogenesis, 2) post-anthesis, and 3) pollen tube growth. Pollen aging produced a progeny population with a greater mean plant size and a lower coefficient of variation than the control progeny. High temperature (29.5 °C) applied both during microsporogenesis and pollen tube growth resulted in progeny populations which were significantly taller and, in one case, had a larger leaf number than the control populations. In contrast, air dried pollen resulted in a progeny population which had significantly smaller character means and larger coefficients of variation than the control population. Also, low temperature (15 °C) during pollen tube growth yielded progeny with reduced branch number and a larger coefficient of variation than the control progeny. In cases where progeny derived from selected microgametophytes were found to differ from the control offspring, corresponding shifts in the reciprocal cross were not observed. For the temperature stress treatments, the lack of reciprocal differences may be related to the different temperature adaptations of the two ecotypes. These results suggest that microgametophytic selection can be effective in shifting the mean of the progeny generation; however, the results obtained will vary depending upon the selection pressure, stage of selection, and the parents used.     

Effects of pollination intensity on the vigor of the sporophytic and gametophytic generation of Cucurbita texana

 We examined the effects of pollen competition (pollen load size) on sporophytic vigor and gametophytic performance in Cucurbita texana, a wild gourd, while controlling for alternative interpretations of the data. Under field conditions we compared the vigor of progeny produced from large and small pollen loads and examined the in vitro performance of the pollen produced by the progeny. We found that the progeny from large pollen loads germinated faster and had a greater reproductive output (male flowers and fruits) than progeny produced from small pollen loads. In addition, we found that the pollen produced on plants derived from large pollen loads grew faster in vitro than the pollen produced on plants derived from small pollen loads. These findings indicate that pollen competition affects the performance of the resulting sporophytic generation and the microgametophytes they produce. Received: 26 January 1997 / Revision accepted: 25 June 1997     

Multiple paternity in common bean (Phaseolus vulgaris L., Fabaceae)

We report on two field experiments that were conducted in 1991 and 1992 at the South Coast Extension and Research Center, Irvine, CA, to study the incidence of multiple paternity in the common bean (Phaseolus vulgaris L.). Hypocotyl color and shikimate dehydrogenase (Skdh) isozymes were used as genetic markers. The white-seeded cultivar ‘Ferry Morse 53’ (FM 53) was used as the female parent. This cultivar is homozygous recessive (pp) for hypocotyl color. The pollen source parents were three homozygous dominant (PP) purple-hypocotyled, black-seeded cultivars. Three cultivars, ‘ICA Pijao,’ G4459, and the maternal parent FM 53, are of Mesoamerican origin and homozygous for the fast (F) allele at the Skdh locus. The other cultivar, Black Valentine, is of Andean origin and is homozygous for the slow (S) allele at the Skdh locus. Overall, 6 125 pods were obtained from 57 and 111 plants harvested individually in 1991 and 1992, respectively. All progeny, 28938 seeds, were scored for hypocotyl color at the seedling stage. The purple-hypocotyled seedlings were genotyped for the Skdh locus to identify their pollen parents. Multiple paternity was identified in all the pods with hybrid seeds (i.e., those of intercultivar crosses) at 5.8% and 8.1% in 1991 and 1992, respectively. All multiply sired pods produced both nonhybrid and hybrid seeds. As many as three successful fathers per pod were identified, but the number of markers limited measuring higher levels of multiple paternity. Most multiply sired pods (≈70%) were filled by nonhybrid seeds plus a single hybrid seed. Ovule position effect within multiply sired pods was inferred from the nonrandom distribution of hybrid seeds within a pod. On average, hybrid seeds occurred more frequently in ovules in position 7 (most basal) and in position 1 (most stylar) than in ovules in the middle positions of the pod.     

The relationship between pollen grain size and progeny gender in dioecious Silene latifolia (Caryophyllaceae)

Summary Experiments and observations conducted during the past 90 years have provided conflicting evidence concerning the existence of a size difference between pollen grains containing an X chromosome (female-determining) and those containing a Y chromosome (maledetermining) in dioecious Silene latifolia. Were such a size difference to exist, this might explain, at least in part, the observation that X-bearing pollen tubes reach the ovary more quickly, on average, than Y-bearing pollen tubes. We tested for such a size difference by separating pollen collected from single anthers into three size classes: small, large, and random. Fruit set (number of pollinated flowers that set fruit) and seed set (number of seeds per capsule) did not differ for these three pollination treatments. Progeny sex ratios resulting from these three pollen size classes also did not differ significantly. Thus, pollen grain size is not affected by which of the two sex chromosomes is present. Our experiment is the first direct test of this relationship. Based on our results, size differences should not be invoked to explain competitive differences in male- and female-determining microgametophytes.     

EFFECT OF OVULE POSITION ON SEED PRODUCTION,SEED WEIGHT,AND PROGENY PERFORMANCE IN PHASEOLUS COCCINEUS L. (LEGUMINOSAE)

We examined the effect of ovule position within the ovary on the probability of seed maturation, on seed weight, and on progeny performance in the outbreeding legume Phaseolus coccineus. Ovaries of P. coccineus possess six linearly arranged ovules (ovule position one = stylar end). We found that in both 1987 and 1988, ovule position had a significant effect on the probability of seed maturation under field conditions. In 1987, ovule positions one. two, and three had a higher probability of maturing seeds than the three most basal ovule positions. In 1988, the probability of producing a mature seed in ovule position one was more similar to the three most basal ovule positions than to ovule positions two and three. The position of the ovule in the ovary had no significant effect on seed weight in 1987, but it had a significant effect in 1988. Overall, seeds from ovule positions one, two, and three tended to produce heavier seeds than the three most basal ovule positions. The effects of ovule position on progeny performance were determined in a greenhouse and a field study. In the greenhouse study, we found no significant overall effect of the position of the ovule that produced the seed on progeny performance. In the field study, we did find a significant ovule position effect on several measures of reproductive performance as well as an overall effect on reproductive performance. In addition, we found a significant interaction between ovule position and number of seeds per fruit. Progeny from the stylar end of the fruit outperformed the progeny from the peduncular end in fruits containing many seeds, whereas there were no significant differences between progeny produced in the stylar and peduncular ends of fruits containing few seeds. Causes of position effects are unknown but hypotheses abound.     

Pollen load size, reproductive success, and progeny kinship of naturally pollinated flowers of the tropical dry forest tree Pachira quinata (Bombacaceae)

Several studies have demonstrated, using controlled pollinations, that the number and identity of pollen grains deposited onto a flower's stigma affect the reproductive success of plants. However, few studies have shown this relationship under conditions of natural pollination. Using the tropical dry forest tree Pachira quinata, we evaluated the relationship between the number of microgametophytes per pistil and the number of sires with respect to the production of fruits and seeds in a natural population of Pachira quinata. Our study demonstrates that fruit and seed production are directly related to the number of microgametophytes per pistil in natural populations of P. quinata. Only 6% of the marked flowers developed into mature fruits and 10% of the marked flowers initiated fruits but later aborted them. A mean of 23 pollen grains were required to produce a seed. Flowers with >400 pollen grains on the stigma always developed into mature fruits, whereas flowers that received <200 grains never matured fruits. Half of the pollen grains transferred to a flower stigma germinated and developed pollen tubes to the base of the style. The number of pollen grains on a stigma explained 34% of the variation in seed number per fruit, and the number of seeds produced per fruit is positively correlated with the size of the seeds. The population of P. quinata studied is predominantly outcrossing, and seeds within fruits are sired by one or a few donors. The total seed crop within trees was sired by three to five donors. Our study examined the implications of the variation in size of microgametophyte loads per pistil with respect to the breeding system and the paternity of progeny under natural conditions. The competitive ability of pollen and pollen tube attrition are important factors regulating fruit production in P. quinata.     

Use of 2n pollen-producing triploid hybrids to introduce tetraploid Mexican wild species germ plasm to cultivated tetraploid potato gene pool

Summary Tetraploid (2n=4x=48) 2EBN Mexican wild species in the series Longipedicellata, which consists of Solanum fendleri, S. hjertingii, S. papita, S. polytrichon, and S. stoloniferum, were crossed with two 2EBN cultivated diploid (2n=2x=24) clones. The resulting triploid hybrids (2n=3x=36) produced 2n pollen (triplandroids) by the mechanism of parallel orientation of anaphase II spindles. The percentage of stainable pollen in 520 triploids ranged between 0 and 23.5%, with a mean of 2.7%. Triploids producing between 13.0 and 23.5% stainable pollen were crossed as staminate parents to the tetraploid cultivars, resulting in abundant pentaploid (2n=5x=60) and near-pentaploid hybrid progeny. Crosses of triploids with lower percentage of stainable pollen as pollen parent to the tetraploid cultivars did not yield fruit, unless rescue pollen from a tetraploid cultivar was added 2 days later. Pentaploid hybrids were selected among selfed tetraploid progenies using morphological and isoxyme markers transmitted from their cultivated diploid parents. These pentaploid hybrids were vigorous and had uniformly sterile pollen. They were female fertile and were crossed with tetraploid cultivars, yielding an average of 19 seeds per fruit. Triplandroids provide the opportunity of transferring 2EBN tetraploid Mexican wild species in the series Longipedicellata germ plasm into the 4EBN cultivated potatoes.Cooperative investigations of the ARS, USDA, and the Washington State University Agricultural Research Center, Prosser, WA 99350, USA. H/LA Paper No. 90-03, College of Agriculture and Home Economies Research Center, Washington State University, Pullman, WA 99164, USA     

NONRANDOM MATING IN WILD RADISH: VARIATION IN POLLEN DONOR SUCCESS AND EFFECTS OF MULTIPLE PATERNITY AMONG ONE- TO SIX-DONOR POLLINATIONS

Nonrandom mating in plants may involve differential success of self- vs. outcross pollen, of pollen from relatives vs. nonrelatives, and of pollen from several compatible donors. Of these, the last is least understood. Wild radish plants were crossed with six pollen donors in all possible combinations to investigate differences in pollen donor performance and to compare multiply sired fruits with different numbers and identities of fathers. Pollen donors differed in number of seeds sired, position of seeds sired within fruits, and mass of seeds sired. Performance in single pollination was not a good predictor of performance in mixed pollination. Performance in two-donor pollinations was a better predictor of the outcome of multidonor pollinations. Multiply sired fruits were selectively filled, but not all multiply sired fruits were equal. Both the number and identity of donors affected fruit filling. These results and others with wild radish suggest that variation in pollen donor ability to sire seeds is relatively common in wild radish. Since wild radish plants in the field typically receive pollen from several donors, there may be considerable opportunity for selection at this stage of the life cycle in wild radish.     

Selective seed abortion affects the performance of the offspring in Bauhinia ungulata

BACKGROUND AND AIMS: Under the microgametophytic competition hypothesis, a non-random pattern of seed abortion is expected, in which only the most vigorous seeds reach maturity. In a previous study, it was found that Bauhinia ungulata (Fabaceae) exhibits a pattern of seed abortion dependent on the position of the ovule within the ovary; ovules located in the stylar half of the fruit, close to the point of entry of pollen tubes to the ovary, have a low probability of seed abortion, whereas ovules in the basal half of the fruit are aborted with a high probability. METHODS: A series of experimental fruits was generated, in which ovules from either the stylar (treatments 1 and 2) or the basal (treatments 3 and 4) half of fruits were destroyed, to evaluate whether these patterns of selective seed abortion have an effect on the vigour of the offspring in B. ungulata. KEY RESULTS: Only 53 % of the seed from control fruits germinated. Seed set in fruits from treatments 1 and 2 showed a significantly lower (33-43 %) percentage of germination; the germination of seeds from fruits in treatments 3 and 4 (49-51 %) did not differ from control seeds. In addition, it was found that the differences in vigour of the offspring are not random with respect to the position of the ovule in the pod. CONCLUSIONS: The overall performance of the seeds correlated with their likelihood of maturation. Seeds located at the basal half of the treatment fruits showed lower values of vigour than seeds located on the stylar half. The differences were more marked for early measures of fitness.     

Effects of soil phosphorus on pollen production,pollen size,pollen phosphorus content,and the ability to sire seeds in Cucurbita pepo (Cucurbitaceae)

To determine the effects of soil phosphorus on pollen production, pollen grain size, phosphate concentration per pollen grain, and the siring ability of pollen, two cultivars of the common zucchini (Cucurbita pepo) were grown under two soil phosphorus conditions in an experimental garden. Overall, soil phosphorus availability had a significant effect on reproductive output through the female function and on traits affecting the male function of plants (staminate flower production, pollen production per flower, and pollen grain size). In addition, pollen produced by plants in the high phosphorus soils had a higher phosphate concentration than pollen produced by plants in the low phosphorus soils. A pollen mixture experiment revealed that pollen produced by plants in the high phosphorus treatment sired significantly more seeds than pollen produced by plants in the low phosphorus treatment. This study showed that growing conditions such as soil phosphorus can influence the size of a pollen grain and its chemical composition, which, in turn, can affect its ability to sire mature seeds.     

Reproductive biology of the dry forest tree Enterolobium cyclocarpum (guanacaste) in Costa Rica: a comparison between trees left in pastures and trees in continuous forest

We compared the rate of pollen deposition, the likelihood of fruit production, the number of seeds per fruit, the outcrossing rate, and the progeny vigor of the tropical dry forest tree Enterolobium cyclocarpum for individuals in pastures vs. individuals in continuous forest. We found that flowers from trees growing in continuous forests were more likely to have pollen deposited on their stigmas than flowers from trees in pastures (52.1 vs. 32.3%, respectively). We also found that trees from continuous forests were almost six times more likely to set fruits and produce more seeds per fruit than trees in pastures. Morever, progeny from trees in continuous forests were, on average, more vigorous than the progeny from trees in pastures, as indicated by 12 of 16 indicators of plant vigor. However, there was no significant difference in the multilocus estimate of the outcrossing rate between the two groups of trees (tm = 1.00 and 0.99 for trees from continuous forest and trees from pastures, respectively). But there are differences in the correlation of paternity between the progeny of the two groups, where the progeny from trees in pastures showed a lower correlation of paternity than progeny of trees from continuous forests (r(p) = 0.104 and r(p) = 0.189, respectively). We argue that the mechanisms that regulate progeny vigor are disrupted in trees from pastures. We discuss the implications of these findings for the conservation of E. cyclocarpum.     

Sequential style removal in highbush blueberry,Vaccinium corymbosum L.: Effects on fertilization success and seed germination

Plants of the blueberry cultivar Bluecrop were pollinated with pollen from the cultivars Sierra and Duke under controlled conditions. After pollination, styles were removed at 6 h intervals from 1.25 days to 4.00 days. An increase in the number of developed seed was seen across time, ranging from 1 seed per fruit (s/f) at 1.25 days, to a peak of 63.4 s/f at 3.25 days in 1992, and from 0.1 s/f at 1.25 days to 56.0 s/f at 3.75 days in 1993. Germination tests of 200-seed seedlots of Bluecrop×Sierra exhibited a significant trend toward earlier germination by seed from earlier style removal times. Seed weight measured in the cross Bluecrop×Duke, however, showed decreasing weights with longer style removal times, suggesting that seed size may time of germination. A multiple regression of seed size and stylar removal time against time of germination showed that style removal time was the only significant source of variation for germination time. Results suggest that pollen vigor, as reflected in the speed of growth down the style, may have a direct bearing on characteristics of the seed and subsequent plant.     

GAMETOPHYTIC SELECTION IN RAPHANUS RAPHANISTRUM: A TEST FOR HERITABLE VARIATION IN POLLEN COMPETITIVE ABILITY

Competition among many microgametophytes for a limited number of ovules can lead to both nonrandom fertilization by pollen genotypes and selection for greater sporophytic vigor. The evolutionary implications of this process depend on the extent of heritable genetic variation for pollen competitive ability. Using flower color in wild radish as a genetic marker, we demonstrate differences among pollen donors in competitive ability. Significant differences were found in four out of five pairs of donors. For three pairs of donors, competitive differences were observed in certain maternal plants but not others. To test for heritability of pollen performance, we conducted a selection experiment. We manipulated the intensity of pollen competition for two generations and then tested for differences in the performance of pollen from two selected lines. Competitive ability of pollen derived from each line was assessed relative to a standard unrelated pollen donor, using pollen mixtures on six wild maternal plants. The intensity of previous pollen competition had no overall effect on the proportion of seeds sired by each selected line. In two maternal plants, pollen from intense previous competition was actually inferior, contrary to expectation. Thus, we found no evidence for heritable variation in this trait. Other factors, such as male-female interactions, may influence the outcome of pollen competition. Prevailing theory on the genetic basis of effects of pollen competition on subsequent generations is not supported by our results. Improved protocols for future experiments are discussed.     

Expression of stylar incompatibility in the Andean clonal tuber crop oca (Oxalis tuberosa Mol., Oxalidaceae)

Long-, mid-, and short-styled clonal accessions of oca (Oxalis tuberosa) were intercrossed in a complete diallelic design. Pollen tube growth in styles was monitored in all diallelic crosses. Pollen fertility was estimated by two tests: staining of pollen grains with aceto-carmine and detection of β-galactosidase activity by the substrate X-Gal. The two methods of pollen fertility estimation were equally useful to detect fertility levels. Pollen originating from short stamens had the highest fertility (85%) and pollen from long stamens had the lowest fertility (70%). Pollen fertility was high throughout, but its degree varied with the stylar morph on which the pollen was formed. Long-styled accessions had the highest rates of fertile pollen. Differences in pollen fertility at different anther levels in the same style morph were also apparent. Pollen grain diameter of the six morph-anther level combinations was inversely correlated with pollen fertility. Pollen grains from long stamens were the largest and pollen grains from short stamens were the smallest. Neither pollen fertility nor pollen grain size had an influence on pollen tube growth in the style or on fruit and seed set. Pollen tubes growing within the styles were inhibited at a different level for each of the 18 cross combinations in the diallel. Although legitimate crosses had greatest pollen tube growth, some of the illegitimate inter- and intramorph crosses had equally high scores. Of all illegitimate crosses, mid-styled seed parents had the lowest level of stylar incompatibility. Fruit and seed set were highly correlated with the extent of pollen tube growth in the style. The number of pollen tubes entering ovules in a flower was in good agreement with the number of seeds produced per fruit. Therefore, it is concluded that stylar incompatibility is the major determinant of limited seed formation in oca even in the most successful legitimate cross combinations. Received: 1 September 1999 / Revision accepted: 4 April 2000     

EFFECTS OF NONRANDOM SEED ABORTION ON PROGENY PERFORMANCE IN PHASEOLUS COCCINEUS L

In this study, we sought to determine if Phaseolus coccineus normally aborts potentially viable seeds, and whether seed abortion is nonrandom with respect to progeny vigor. The ovaries of Phaseolus coccineus typically have six linearly arranged ovules. The three ovule positions at the stylar end are more likely to mature seeds than the three ovules at the basal end of the ovary. When we destroyed the developing ovules at the stylar end of the fruit after fertilization but before seed abortion, there was a significant increase in the probability that the ovules in the three basal positions would produce a mature seed. The probability of seed maturation in control fruits (no ovules destroyed) ranged from 38.3 to 42.7% over the three field seasons, whereas in the experimental fruits it ranged from 64.3 to 79.7% (similar to that of ovules at the stylar end in control fruits). We did not find any significant change in the probability of seed production in the three ovule positions in the stylar end of the ovary (the positions with high probability of seed maturation) when the basal ovules were experimentally destroyed. These results indicate that potentially viable seeds are regularly aborted in P. coccineus, with seed abortion more frequent in the basal ovule positions than at the stylar positions. In two greenhouse studies and one field study, we compared the vigor of progeny produced in the control fruits with the vigor of progeny produced in fruits where three ovules (either stylar or basal end) were destroyed. We found that the performance of the progeny from the three positions at the stylar end of the control fruits did not differ from the performance of the progeny from experimental fruits in which the three basal ovule positions were destroyed. In contrast, the progeny from the basal positions of the control fruits outperformed the progeny from the experimental fruits when the seeds were produced in the three basal ovules (stylar ovules destroyed). Our findings indicate that when the experimental treatment increased the probability of seed maturation, there was a significant decrease in the average vigor of the progeny in the remaining (basal) ovule positions.     

Inbreeding depression affects pollen performance in Cucurbita texana

Most studies of the effects of inbreeding focus on vegetative vigor and reproductive output through the female (fruit and seed) function. This study not only examines the effects of inbreeding on the female function but it also examines the effects of inbreeding on pollen performance both in vitro and in vivo. This study used Cucurbita texana, a wild gourd, and was performed under field conditions. In vivo performance was assessed by placing equal amounts of pollen from either the inbred or outcrossed plants onto a stigma together with pollen from a tester line. As with other studies, we found that outcrossed plants had greater reproductive output (male flowers and fruits) than plants produced from self pollinations. Unlike most studies of inbreeding depression, which mostly ignore the male function of plants, we also found that the pollen produced by outcrossed plants had faster pollen tube growth in vitro than the pollen produced on selfed plants. Moreover, pollen from selfed plants sired significantly fewer seeds than pollen from outcrossed plants under conditions of pollen competition (i.e. the number of pollen grains deposited onto the stigmas was larger than the number of available ovules). These findings indicate that inbreeding affects the performance of the resulting sporophytic generation and the microgametophytes they produce.     

Onion pollination by blowflies and honeybees in large cages

Three experiments to investigate the effect of pollination by blowflies (Calliphora and Lucilia spp.) and honeybees (Apis mellifera) on seed yield and crossing level were made using a mixture of plants from two open-pollinated cultivars of onions (Allium cepa) in breeders' cages. No consistent differences were found between pollinators. In two experiments, fly pollination gave significantly higher yields than bee pollination but in the third, bee pollination gave slightly more seed. Crossing between cultivars, assessed by progeny testing on one cultivar, varied between 22-7 and 55-9%. Differences in crossing level attributable to the pollinators were not consistent and were not or only just statistically significant, but there were significant and substantial differences between the experiments themselves. Where more umbels were produced by the red-bulbed cultivar than by the yellow, higher crossing levels were found in the yellow cultivar. The highest crossing levels were found in the only experiment where insects were not introduced until peak flowering time, but where seed yield was comparatively low. Crossing levels were lowest where the umbel numbers were approximately equal. Individual umbels studied showed that while seed set was usually heaviest at and just before the mid-flowering date, this period could correspond with the lowest levels of crossing during the flowering period, particularly if flowering of the two cultivars was not completely synchronous. In field plots, approximately 75% of crossed seeds sown survived to give harvestable bulbs, but fewer than 50% of selfed seeds did so. In an experiment in which crossing levels from different pollinators were similar but survival levels differed significantly, it was deduced that bees performed more crossing within a cultivar than flies.     

Characteristics of self-incompatibility in Schlumbergera truncata and S. x buckleyi (Cactaceae)

The influence of self-incompatibility (SI) on fruit set, seed set, and pollen tube growth was investigated in Schlumbergera truncata (Haworth) Moran and S.xbuckleyi (T. Moore) Tjaden. Four Schlumbergera clones were crossed in a complete diallel to verify the presence of SI. Fruit did not set when the clones were selfed or when two of the clones were crossed reciprocally, but all other outcrosses yielded fruit which contained 100–200 seeds each. Compatible outcrosses were characterized by large numbers of pollen tubes in the style and ovary cavity at 72 h after pollination. When pistils were selfed or incompatibly crossed, pollen tubes were inhibited in the upper third of the style and few pollen tubes reached the base of the style by 72 h after pollination. Schlumbergera exhibits several characteristics often associated with sporophytic SI systems (tricellular pollen and dry stigmas with elongate papillae), together with those commonly observed in gametophytic SI systems (stylar inhibition of incompatible pollen tubes and absence of reciprocal differences in outcrosses).Publication no. 3174 of the Massachusetts Agricultural Experiment Station